Steve Hayden's Miscellaneous Knowledge Page
(Includes reference for non Windows
2000 platforms)
The following sections are for reference only
Miscellaneous
Platform Installation and Configuration
Installing
Visualage for Java (VAJ) and the WebSphere Test Environment on NT
Installing AIX from Scratch
Installing
NT from Scratch (Under Construction)
Installing Linux
(RedHat) from Scratch
Instructions to
Setup Mainframe Access
Instructions
to Setup PC to Workstation Access
Installing DB2
Universal Database on AIX
Installing DB2
Universal Database on NT
Netscape
Enterprise and IPlanet Server Installation
Netscape Browser Installation
C/C++ Installation on AIX
C/C++
Installation and Compiling/Building on NT
WebSphere CLASSPATHs
Building and Compiling
Debugging
Utility (DBX) on Solaris and AIX
Tracing and Logging
WebSphere Database Cleaning
Setting Up WS for
Persistance
Setting
Up WebSphere With *.jsp's and beans.
Setting
Up Multiple Application Servers To Share Session Data
Options for
Servlet Access Within WebSphere
Creating Model/Clones
Using OSE Remote
AKStress Installation
Unix/Linux/AIX
Operating System Commands/Utilities
Miscellaneous Notes/Tidbits
Author’s Notes
Miscellaneous
Platform Installation and Configuration
WebSphere
3.5 Operating Environment Prerequisites
WebSphere
3.0x Installation on AIX
WebSphere
Installation from CD
WebSphere
3.5x Installation From DFS Build
Administration
Server Startup
Default
Host And Default Server Configuration for Netscape Instance
Modifying
and Exporting VAJ Code for AIX
WebSphere
Install On NT
JDK
1.1.8 Installation
JSDK
Installation
DB2
6.1 and Associated Fixpak Installation
WebSphere
and IBM HTTP Server Installation
Visualage
for Java (VAJ)
Running
A Servlet In The WebSphere Test Environment
Changing
WebSphere Code In WebSphere Test Environment in Visualage for Java
Installing
Fixpacks on NT
WebSphere 3.5 Operating Environment Prerequisites
Operating Systems:
NT4 SP4/SP5/SP6a
AIX4.3.3.02
Solaris 2.6, Solaris 2.7
HP-UX 11.0
Supported JDKs
JDK 1.2.2
IBM JDK 1.2.2 FP 6 on NT/Windows 2000
IBM JDK 1.2.2 FP6 on AIX
Solaris JDK 1.2.2 FP5a on Solaris
HP JDK 1.2.2 FP4 + 3 patches on HP-UX
JDK is bundled with the product install
Supported Web Servers
Apache 1.3.12
IBM HTTP Server 1.3.12
Microsoft IIS 4.0 and 5.0
Netscape Enterprise Server 3.51/3.63
No HP-UX Support
iPlanet Enterprise Server 4.01
Lotus Domino 5.0.2b
Dropped Lotus Domino Go 4.6.2.5
Supported Databases
DB/2 UDB 6.1 FP4
Oracle 8i (8.1.6)
InstantDB 3.1.3
Sybase Adaptive Server Enterprise 12.0
AIX and Solaris only
Dropped DB/2 5.2
Dropped Oracle 8.0.5
WebSphere
3.0x Installation on AIX
Steve, start cleaning this!
Note: Both of the two following examples assume the Netscape web server will be used and has been installed. You may choose to install the IBM HTTP Server and associated plug-in if you wish. Instructions on the installation of Netscape are in later sections. (WebSphere 3.02 might require at least AIX 4.3.3 with fixpack 4)
WebSphere Installation from CD
At the command prompt, export the following environment variables (or alternatively include in your ".profile"):
export JAVA_HOME=/usr/jdk_base
export NS36=/usr/ns363/es363/https-level3/config (or whatever
server you want configured)
Using the mounted WebSphere Advanced Version CD, change directories to "/CD/aix" and enter the command "./install.sh" at the command line. This will invoke the WebSphere installion GUI. You will first be presented a window which states this is the installation program, click "Next". You will then be presented a window which displays your JDK version (JDK should have been automatically installed during your AIX 4.3.3 installation). Click "Next". The next window will be to choose the type of installation. As you need to install for Netscape web server, select "Custom Installation" and click "Next".
The "Production Application Server", "Administrator's Console", and "Documentation" are already selected on the left side of the window, and "Core Server" is already selected on the right side. First select the "Netscape V3.6 plug-in", or alternatively the "IBM HTTP Server plug-in" if the IBM HTTP Server is installed. If not and you want to install it you can also select the "IBM HTTP Server 1.3.x" to be installed. Also select the "Install a default configuration" option and click "Next". If you have selected the option to install the Netscape Plug-in, you will then be prompted for your Netscape configuration file (i.e. "/usr/ns363/es363/https-level3/config/obj.conf").
In the next window in the "Security" section, enter a User id and associated password (i.e. root/id10t) if you wish to install WebSphere enabling security. In the "Database" section, select "DB2", enter the name of the database WebSphere will use (i.e. "was"), and in the "Userid" and "Password", enter "root/id10t" (note, root was used as an example, but it works). Then enter the "DBHome" (i.e. /home/db2inst1"). The "DB URL" will automatically be filled in, click "Next". You will probably get the next screen which tells you prerequisites have been exceeded. Click "Next". You might then be prompted for your Netscape configuration file (i.e. "/usr/ns363/es363/https-level3/config/obj.conf"). You will then see a pop-up "Do you want to begin copying files, click "OK", and wait a month.
In about a month, you will see a window "finished installation" and asking if you want to see the "readme" file. This is up to you.
Go to the section "Administration Server Startup" for additional instructions
on configuring WebSphere.
WebSphere 3.5x Installation From DFS Build
Using the build retrieved from DFS, change directories to where you have untarred the build and enter the command "./install.sh" at the command line. This will invoke the WebSphere installion GUI. You will first be presented a window which states this is the installation setup program, click "Next". The next window will be to choose the type of installation. As you need to install for Netscape web server, select "Custom Installation" and click "Next".
The "Application and Administrative Server", "Administrative Console" and "Webserver Plugins" will already be selected, select "Samples" also and click "Next". In the next screen select the appropriate plugin for your configuration and click "Next".
If you have selected the option to install the Netscape Plug-in, you will then be prompted for your Netscape configuration file (i.e. "/usr/ns363/es363/https-level3/config/obj.conf").
You will then be queried for your database information. The database name is "was". Then enter the "DBHome" (i.e. "/home/db2inst1"). The "DB URL" will automatically be filled in. Use the root id and password, which you must enter twice, then click "Next".
In the next window in the "Security" section, enter the root id and associated password if you wish to install WebSphere enabling security. Use the remaining defaults and click "Next".
You will then be queried for the "Destination Directory", use the default and click "Next". The options you selected will be displayed, click "Next".
You will then see a pop-up "Do you want to begin copying files, click "OK", and wait a month.
In about a month, you will see a window "finished installation" and asking if you want to see the "readme" file. This is up to you.
Go to the section "Administration Server Startup" for additional instructions
on configuring WebSphere.
In the /usr/WebSphere/AppServer/bin directory, modify the file "admin.config"
by changing the value of "install.initial.config" from "false" to "true"
(this will already be done if this is a new install and you selected the
option to install initial configuration/samples). Now invoke the
server startup shell (i.e. "./startupServer &"). This will take several
minutes to complete. During this startup logs are being written to the
"tracefile" in the directory "/usr/WebSphere/AppServer/logs" "WebSphere
Administration server open for e-business" appears in the tracefile log,
the admin server has been successfully started. You may then invoke the
command "./adminclient.sh &" to start the GUI.
Default Host And Default Server Configuration for Netscape Instance
From the Administration Console GUI, select to WepSphere Administrative Domain, and scope down to the "default_host". In the default host "Advanced" tab, add your server port number to the "Host Alias" entries for your server (this will be the domain name and ip address entries, i.e. "id10t:20001" and "9.27.226.102:20001").
From the topology view, scope down in the "WebSphere Administrative Domain" to your machine/node (i.e. "id10t"), Select the "Default Server", and while on the "Default Server" field, right mouse click and select "Start". (IT IS STRONGLY RECOMMENDED YOU STOP YOUR DEFAULT SERVER BEFORE YOU MAKE ANY CHANGES TO IT’S CONFIGURATION!)
To stop the admin server, from the topology view, select your machine (i.e. "id10t"), and with the right mouse button, select "Stop". You will get a warning stating everything will stop.
If this does not work you will need to kill all associated processes,
and hope the thing is not broke when you try to restart everything.
Options for Servlet Access Within WebSphere
This section describes a way to install and uninstall WebSphere with minimum disruption to your personal servlets. When you install WebSphere with "Install a default configuration" selected, a set of example servlets are installed. If you wish to create your own servlets and have WebSphere administer them also, you can add them to your environment by using the Admin GUI as desired.
Alternatively you can create a (sub)directory/folder on your machine
in which you servlets will reside (i.e. "/usr/HTTPServer/htdocs/en_us/servlets").
You then need to add this path to your application. This is done at the
Administration Console by selecting the appropriate web application, and
in the "Advanced" tab, add the path your servlets reside in to the "Classpath",
"Applying" it, and restarting your application. At this point you can go
to a browser and invoke that servlet (i.e. "http://id10t/servlet/YourServletName").
Steve, think about the following.
Obtain the JDK environment from the Austin ftp site "zircon.austin.ibm.com"
using your DAAT logon, and go to the "/:/projects/boss/dev_tools/java/nt"
directory and retrieve the file "ibm-jdk-n118p-win32-x86.exe", placing
it on your local disk drive. From Windows NT Explorer, double click on
this file and use all defaults for this installation.
The JSDK contains the classes required for compilation of servlets on your machine. The WebSphere installation includes a "servlet.class" file which contains these classes. However, you may elect to install the JSDK from sun (described later in this section. Regardless of which classes you choose, you must modify your machines CLASSPATH to include the directory path of the appropriate jar file. In the WebSphere directory structure this will be found in your "lib" directory (i.e. "/WebSphere/Appserver/lib". Include the appropriate path in your CLASSPATH. (i.e., chose "Start", "Settings", "Control Panel", and in the resultant pop-up window, double click on "System". Add the path of your jar file to the "CLASSPATH" System Variable.
The Java Servlet Development Kit can be acquired at http://java.sun.com/products/servlet/download.html . Download the file, execute it, and it will create a "Jsdk2.0" directory on your hard drive. You must then add the pathname of the "jsdk.jar" file to your system classpath. From "Start", "Settings", "Control Panels", and double click on "System".
In the System Properties pop-up, select the "Environment" tab, select
"CLASSPATH", and add the path to your jsdk jar file to the end (i.e. "…..;C:\Jsdk2.0\lib\jsdk.jar").
You can then go to a DOS prompt, go to the directory your servlet java
file exists, and use the javac command to compile your servlet (i.e. "javac
MySillyServlet.java"). This will generate a "MySillyServlet.class"
file which is an executable servlet.
(VisualAge should have already modified your classpath to include the
normal JDK libraries).
DB2 6.1 and Associated Fixpak Installation
Stick in the DB2 6.1 CD, wait a few seconds, follow the directions and choose defaults where possible.
Select CUSTOM install, select command center and control center.
When queried for a Security user id and password, use an id other than the Administrator’s, and for the database user and password, use this same id/password (it is preferable to use an id that is the same as the machine name.
Stick in the DB2 6.1 Fixpack 2 CD, wait a few seconds, follow the directions and choose defaults where possible. If queried for a password, enter in the same userid/password as above.
Upon installation of WebSphere (described later), the installation will tell you it has run the script to create and configure the Websphere Application Server database ("was"), however, it lied to me. After db2 is installed you will be required to restart your machine, upon which, db2 will be started.
Either from Windows NT Explorer or from a dos prompt, execute the script "createdb2.bat" which resides in the directory "\WebSPhere\Appserver\bin". This will create the was database.
At this point you will need to change the heap size. From the "Start", "Programs" "DB2 for Windows NT", invoke the "Command Center". Type the commands "connect to was", in the "Script" tab, and then using the mouse, select "Script" from the top menu bar and select "Execute" to execute the command. Then, also from the script tab, type "update db config for was using applheapsz 256", select "Script" from the top menu bar and select "Execute" to execute the command.
You can use the "Start", "Programs", "DB2 for Windows NT", and invoke
the "Control Center" to monitor and review your was database.
WebSphere and IBM HTTP Server Installation
(Note: the author had to install these in 2 discrete steps, first doing a custom install and only installing the IBM HTTP Server, then did a normal (custom) WebSphere install).
Obtain a WebSphere 3.02 CD and insert it into your CD drive. After several seconds you will see a pop-up quering you for your choice of language, select your choice and click "OK", and in the dialog screen telling you to stop everything else because you are going to install WebSphere, click "Next". Select "Custom Installation"and click "Next".
Select all components including the "Configure administrative domain with default application server and a default application", the "IBM HTTP V1.3.6" server (if desired/required), and select the "IBM HTTP Server V1.3.6" plugin (or whatever plugin you desire/require), then click "Next".
The installer will then check what versions of the JDK environment you have, and display your alternatives, select "Other". You should have already installed the JDK 1.1.8 version available from the Austin ftp site, and if this is not selected, you will be allowed to "browse" to the appropriate folder for this version, then click "Next" (you will probably get a warning pop-up stating you have exceeded the requirements, click "OK").
The next screen will be your DB2 information. Enter the same password you entered previously in the database user section in the Security section, and in the "Database" section, this same id/password and the "was" database type, name, path, and url should already be filled in. Click "Next". Accept the default program folder name and click "Next". Sit back and watch the blinking lights (you might be qureied for the location of the configuration files for your web server(s)).
After you have installed WebSphere, you must then restart your computer, and then manually start WebSphere. From your menu bar, "Start", "Settings", "Control Panels", double click on "Services". Select the "IBM WS AdminServer" and "Start" it.
You must then start your IBM HTTP web server using "Start", "Programs",
"IBM HTTP Server", and selecting "Start HTTP Server". You can then
use a browser to connect to your server using your url (i.e. http://shayden/servlet/snoop.
Installing
Visualage for Java (VAJ) and the WebSphere Test Environment on NT
Visualage for Java (VAJ)
Running
A Servlet In The WebSphere Test Environment
Changing
WebSphere Code In WebSphere Test Environment in Visualage for Java
Modifying
and Exporting VAJ Code for AIX Environment
From the ftp site zircon.Austin.ibm.com, log on using your DAAT logon
id and password. In the directory "/:/projects/boss/dev_tools/vaj/nt/vaj3.0.2_GM",
download file "win_daily_entp-3_02_00301w-gold.zip". After ftp’ing,
xtract the zip, and in the directory you extracted the files, run the "Setup.exe"
and follow the instructions (use defaults where possible). Start VAJ, and
if you are queried for a repository owner in a popup, select Administrator
and enter your Administrator password. Afterward, in the "Welcome to Visualage"
popup, click the "Close" button.
Running A Servlet In The WebSphere Test Environment
Note: there are often multiple ways you can invoke actions in VAJ, i.e from the right mouse, control characters, from the menu, etc. In the following sections examples of all are described.
From the VAJ Workbench, select "File", "Quickstart", and in the "Quick Start" popup, select "Features". In the "Features" option, select "Add Feature". Click "OK". In the "Selection Required" popup, select "IBM WebSphere Test Environment" (WTE). Use this same procedure to add the "IBM EJB Development Environment".
From the VAJ Workbench, click the "Debugger" option (the "bug", which is the second icon from the right in the menu bar), or alternatively, select "Workspace", "Tools", and "Launch WebSphere Test Environment". After a few seconds a WebSphere Test Environment pop-up will appear, as well as a console window. The console window will log activity during execution of the WTE.
At this point you can use your browser to connect to the WTE by using port 8080 (i.e. "http://localhost:8080/servlet/HelloWorldServlet"). An example servlet "TestServlet" also exists which is somewhat similar to the "SnoopServlet" which comes with WebSphere.
As an example of changing code, go to VAJ Workbench, select "Projects" tab, scope down from "IBM WebSphere Test Environment", "Default package for IBM WebSphere Test Environment", and double click on "HelloWorldServlet". A new window will pop-up with the members "message" (which is the HTML message displayed in the browser), "destroy" and "init" (both of which simply output a message to the console window), and the "service" method (which actually displays the message). Select the "message" member and change the message displayed with an eye catcher for you. (you can use "ctrl/s" to save your changes or by simply selecting another method the one you are working on will be automatically saved).
Re-execute the servlet from your brower and you should see your changes
in the result.
Changing WebSphere Code In WebSphere Test Environment in Visualage for Java
At this point you can only view some of the WebSphere classes, and not the actual java code in VAJ, as well as, the WTE only contains a subset of the WebSphere java code.
You should now import the source code you wish to change into VAJ. You must first get the java source code for your class(s), either from CMVC or from the Austin build site. Caution must be exercised here in terms of compatibility between the different WebSphere versions of classes you wish to import into VAJ. In addition, there is the likelihood that you will need to import additional code into Visualage, either from the "ncf.zip", "kansas.zip", or "prereqs.zip", each of which exists in the directory where the build you downloaded resides. The task of determining which classes need to be imported, and where they are, involves some creative analysis and decision making. You have the options of either importing the source code, or perhaps the appropriate jar file, or the compiled classes themselves. As well as, in the installed WebSphere's "lib" directory resides the WebSphere jar files, which can be "unjarred" (i.e. "jar -xvof ejs.jar") to obtain the appropriate class(es) for importing.
Back in the Workbench, select the "IBM WebSphere Test Environment", right mouse and select "manage" and "Create open edition".
In this example, we will select the "Session Tracking" code to import. From the Workbench window, in the Projects tab, select "IBM WebSphere Test Environment", and scope down into the "IBM WebSphere Test Environment" package. Select the "com.ibm.servlet.personalization.sessiontracking", right mouse and select "Delete". Select the "IBM WebSphere Test Environment", and right mouse, select "Import", and in the Import popup, select "Directory", and click "Next". You can then browse for the directory your code is in (i.e. "~/ncf/src/java/com/ibm/servlet/personalization/sessiontracking") and click "OK". You will only want the Java files to be imported, so select ".java". Note that as the files are being imported they are being compiled, so you will perhaps see "problems found" that (hopefully) will go away when subsequent classes are loaded. For the session tracking code to be imported, the author had to import the "db" from databeans.jar file from C:\WebSphere\AppServer\lib, and had to import the StaleConnectionException.java class from C:\ws302.sources\kansas\kansas\src\jmon\com\ibm\ejs\dbm\jdbcext
Good luck!
Modifying and Exporting VAJ Code for AIX Environment
Note: this technique will probably work for NT, Solaris, and the rest of the platforms, but as the author only tried it on AIX, ....
Import and modify the appropriate code in VAJ. Select the class you wish to export, and with the right mouse, select "Export". In the resulting "Smart Guide" "Export" pop-up, in the "Select an export destination", select "Jar File" and click "Next". In the next panel, "Export to a JAR file", enter the directory/filename you wish to export to, and select ".class". This will create a jar file which you must ftp to your destination machine.
At this point you should stop your admin server.
After ftp'ing the jar file to your AIX box, you can place the jar in
whatever directory you desire, and then modify the classpath in your "~/WebSphere/AppServer/bin/admin.config"
file by placing that directory path and associated jar file at the beginning
of your classpath specification. Restart your server and watch the fireworks.
Installing Fixpacks (PTF's) on NT
Although you normally would only install the desired fixpack on your
existing version of WebSphere, for this exercise, the author uninstalled
the existing version of WebSphere ( [Start], [Programs], [IBM WebSphere],
[Application Server 3.0], and clicked on "Uninstall WebSphere Application
Server"), then restarted the machine. Upon reboot, deleted the WebSphere
folder, and stopped the HTTP server from the "Control Panel" "Services"
window. Then "dropped" the WAS database ( [Start], [Programs], [DB2 for
Windows NT], and clicked on "Control Center". In the logon pop-up,
enter your DB2 ID/Password and press enter. In the resulting "DB2 Control
Center" window, scope down to your machine, "Instances", "DB2", "Databases",
and select WAS (or whatever your was database name is), right mouse click
and click on "Disconnect", then right mouse click and click on "Drop".
( Now you're ready to start fresh.)
Installing the Fixpack
If you have not done so already, follow the instructions in the section
WebSphere
and IBM HTTP Server Installation
From the site "zircon.austin.ibm.com", logon using your DCE id/password,
and from the directory
"/:/projects/boss/botp/build/(WAS_version)/test/(PTF_version)/nt/zipper"
(where the author installed WAS_version of ASV302", and PTF_version "ptf0015.03"),
select binary ftp mode, and retrieve the appropriate advanced or standard
edition fixpack (i.e. "fixpack_Adv_NT.zip"). From the NT Explorer, scope
to this file and extract the files from the zip file.
Stop all WebSphere applications, the admin server, and all web servers on your machine. Assure the JDK "classes.zip" file (i.e. "c:\jdk1.1.8\lib\classes.zip") is included in your system CLASSPATH. From the DOS command prompt, go to the directory you extracted the fixpack in and type "install", which will install all necessary files.
Restart the admin server from the Control Panel "Services" window, as
well as restart the web server(s). From the [Start], [Programs],
[IBM WebSphere], [Application Server 3.0], click on "Administrator's Console",
and start the default application. You are now in e-bitness.
Installing NT from Scratch (Under Construction)
1) Make a set of NT boot disks. Obtain the NT OS CD and create a set of boot diskettes, and from the CD’s /i386 directory, execute the command "winnt32 /ox" (you will be instructed to put diskettes in at the appropriate time, don’t forget to label them)
2) Boot up with this disk set (you will be instructed to put diskettes in at the appropriate time), and follow the instructions. In the Windows NT Workstation Setup screens, you must delete any exiting partitions and create new ones (C=500meg, or whatever you want) Later you can use the disk administrator to create/partition as desired.335-3356
3) System will boot up from CD and ask you several questions, your Serial
Number, etc. Select defaults except for Net BUI
Installing Linux (RedHat) from Scratch
Additional
Miscellaneous Linux Stuff
Installing
DB2 on Linux
Linux
Java Development Kit Installation
WebSphere
Application Server Installation on Linux
Configuring
Linux for Compiling and Building
WebSphere
Linux Build Environment
Linux Installation Instructions
Format a diskette, go to dos prompt and "format a:". You will be prompted to insert a diskette.
From the Red Hat CD, create a boot diskette. Go to the "dosutils" directory and execute "rawrite.exe". A pop-up will appear requesting you enter the disk image for the boot diskette (i.e. "d:/images/boot.img"), as well as a target diskette drive (i.e. "a:").
Reboot your system with the Red Hat boot floppy in the floppy drive and the Red Hat CD in the CD drive. You will get a "Welcome to Red Hat Linux 6.2" window. Choose the first (default) option (press enter).
Enter the language you wish to use (swaheelei). You will then get a screen which queries you for Model/layout/Deadkeys, which reflects Linux can use these as defaults (Model = "Generic 101-key PC", Layout = "U.S.English w/ISO9995-3", Dead Keys = "Enable dead keys"), press "Next".
You will then get a screen which queries you for Mouse Configuration ("3 Button Mouse (PS/2)"), press "Next". You will see the "Redhat Linux" screen, click "Next". At the next screen you must then select the Install Type ("Custom"), and click "Next", you will see the "Redhat Linux" screen, click "Next".
The next screen is Disk Druid. This is where you create the root directories. First "Delete" the "<not set Mount Point. Click the "Add" button and add the "Mount points" you require. There is a pull-down on the "Mount Point" field for selecting the normal Linux directories. You have various options you can select for the size/type, etc. For example, select the "/" Mount Point, "Size" = 1000 meg, leave "Grow to full disk" as default (unslected), and Partition Type is "Linux native", allowable drives: = "hda", click OK. Other values are "/swap" = 1000meg, (note, this is required, use the same instructions for adding a mount point, but the Partition Type must be "Linux swap"), "/boot" = 1000 meg, "/usr" 2500meg, "/home" = 2000meg, "/var" = 2000meg, "/opt" = 2000meg, "/tmp" = 1000meg, "/usr/local" = 1000meg (for this one select "Grow to full disk"), "/export" = 2000meg. Click "Next".
The next window "Chooose partitions to format", all partitions should already be selected. Select "Check for bad blocks while formatting", and click "Next". The next window "Lilo Configuration", unselect "Create boot disk" (unless you want one), keep remaining defaults and click "Next".
The next window "Network Configuration", unselect "Configure using DHCP", and select "Activate on boot". Some subsequent values will be filled in when you enter values into some of the fields, so assure your IP address, Netmask, and the Network, Broadcast, Hostname, Gateway, Primary/Secondary are entered and correct (you should have already gottn your domain name and ip address from the Austin web site). Click "Next".
In the next window select your time zone.
In the next window Account Configuration, enter your root password and confirmation. You can also create any user accounts from this window.
In the next window "Authentification Configuration", this is nothing special, so click "Next".
The next window "Package Group Selection", go to the bottom and select "Everything". Click "Next"
In the next window "Monitor Configuration", select "IBM 6554 P70 (673, 674, & 6E3)", and click "Next".
In the next window "X Configuration", select "Use Graphical Login" at the bottom of the screen, and click "Next".
The next screen "About to install", click "Next", and wait a month. After the install is complete you will be prompted to press "Return" to boot the machine.
Once the machine has come back to life, you must set the domain name.
Using the "/bin/linuxconf" utility, select the "Name Server specification
(DNS)" option, and in the "Resolver configuration", change the "default
domain" to "raleigh.ibm.com", and click "Accept". Then click "AcctChanges"
button on the bottom of the linuxconf window for your changes to be actuated.
You will then see a panel which states the status of the system needs synchronization.
Click the "Activate the changes" button. (When you quit out of the linuxconf
utility you might be queried to activate the changes again.)
Additional Miscellaneous Linux Stuff
Linux uses a "Red Hat Package Manager" (RPM) utility for installing and uninstalling software. There is a utility "gnorpm" (in "/usr/bin") which is a GUI used for (un)installing, which you may use for software maintenance, however all actions you can perform using the GUI can also be done at the command line. Some common commands (substitute your package name for "IBMWebAS"):
rpm -q IBMWebAS
(shows the rpm package name)
rpm -qi IBMWebAS
(shows lotsa details about the package, location, description, etc)
rpm --uninstall IBMWebAS
(uninstalls the package)
man rpm
(bores you to death with details and "the philosophy" of rpm)
System/User administration is done by the "/bin/linuxconf" utility.
The Apache server comes pre-installed in Linux. To start, go to "/usr/rc.d/init.d" and execute the command "httpd start", to stop issue the command "httpd stop" (no quotes). Other options for the httpd utility are "start|stop|restart|reload|status".
Change to "/etc/cron.daily" directory and execute "./slocate.cron" which sets up the "locate" utility. Then "hostname lucy4.raleigh.ibm.com" to set up the domain configuration.
In the lower left hand corner of the screen is the "Start Menu" icon.
To mount a cd, insert the cd and shortly afterwards you will see a "Question" pop-up, asking if you wish to run "/mnt/cdrom/autorun". Clicking "Yes" mounts the cd. To unmount, select cd icon on the desktop, right mouse, and select "Unmount device".
Netscape browser cannot be run as root.
"pico" is an editor.
To start the apache http server, go to "/etc/httpd/conf" and modified the file "httpd.conf", modified the parameter "ServerName" to a value of "lucy4.raleigh.ibm.com". Then go to "/etc/rc.d/init.d" and execute "./httpd start" to start the Apache server. THIS MUST BE DONE BEFORE INSTALLING WEBSPHERE!
To reboot a Linux machine, go to a terminal window on the system and type "reboot".
With SuSe Linux, with a generic WebSphere install with Apache web server,
you must manually modify the "httpd.conf" file which resides in the "/etc/httpd"
directory. This includes moving the line "LoadModule app_server_module
/opt/IBMWebAS/bin/mod_app_server.so" and "AddModule mod_app_server.c" to
the end of the file. You must also create 2 symbolic links, by first
creating a subdirectory "conf" in the directory "/etc/httpd". Change to
this directory and in it create 2 symbolic links, "srm.conf" and "httpd.conf"
linking from the "/etc/httpd/conf" directory to the associated file in
the "/etc/httpd" directory (i.e. from "/etc/httpd/conf" directory, enter
the command "ln -s /etc/httpd/httpd.conf" (no quotes) and the same for
the "smr.conf" file).
Downloaded db2 from "http://arathorn.raleigh.ibm.com/builds/db2/DB2_FULL_INSTALL_WITH_FIXPAK5"
the file
"DB2_Full_withFP5.tar" (109905920 Bytes). Userid = "test"
Password = "testteam" (no quotes).
Untarred the "DB2_Full_withFP5.tar", changed to the "s000729_EE" directory,
and ran the utility "./db2setup". In
the "Install DB2 V6.1" window, selected the "DB2 Administration Client"
(by pressing "Enter"), tabbed to the "DB2 UDB
Enterprise Edition" and slected it also (by pressing "Enter").
Tabbed to the "OK" button and pressed "Enter".
In the "Create DB2 Services" screen, tabbed to the "Create a DB2 Instance."
and pressed "Enter", which took me to the
"DB2 Instance" screen In the "DB2 Instance" window, kept defaults,
tabbed to "OK", and pressed "Enter". A "(!) A system-generated password,
ibmdb2, will be used." pop-up was displayed, press "Enter". The "Fenced
User"window displayed, I kept defaults, tabbed to "OK", and pressed "Enter".
A "(!) A
system-generated password, ibmdb2, will be used." pop-up was displayed,
press "Enter" to acknowledge.
Returned to "Create DB2 Services" window, tabbed to "( ) Create the
Administration Server.", and pressed "Enter". In
the "Administration Server" screen, accepted defaults and tabbed to
"OK" and pressed "Enter". The notice "(!) A system-generated
password, ibmdb2, will be used." pop-up was displayed. Pressed "Enter".
A "Notice" appeared stating "(!) DB2SYSTEM will be set to 'lucy4'. ", then
pressed "OK"
Returned to "Create DB2 Services" window, tabbed to "OK" and pressed "Enter". The "DB2 Installer" "Summary Report" was displayed, pressed "Enter" to continue. A "Warning" pop-up stating last chance popped up, pressed "Enter".
Go away for a month unless you like sitting on the edge of your seat!
For WebSphere installation you will need to either change the password of the "db2inst1" user in the WebSphere "admin.config" file (i.e., "com.ibm.ejs.sm.adminServer.dbPassword=db2inst1"), or assure the db2inst1 user's password is the default "db2inst1". Changing the password can be done by changing to the user and from the command line executing the "passwd" command, or using the "linuxconf" utility user password changing facility.
Once you have db2 successfully alive, all users which need to use db2
must either execute the command line
". /home/db2inst1/sqllib/db2profile", or otherwise have this
in their ".profile" or shell script they execute when logging on.
One of the attempts was not successful, got "Completed With Errors" report. Upon reviewing log, the "Create DB2 Instance" failed.
DB2 Instance Creation
FAILURE
Create new entry in /etc/services
CANCEL
Update DBM configuration file for TCP/IP
CANCEL
Auto start DB2 Instance
CANCEL
Start DB2 Instance
CANCEL
To correct this, first run the "db2_deinstall" utility to remove the
db2 subsystem. You must then remove the following if they were successfully
created: remove all three db2 users, remove the "/var/db2" stuff,
remove the "/usr/IBMdb2" stuff and remove the "/tmp/db2*" stuff, and kill
all db2 processes running. Re-run "./db2setup".
Linux Java Development Kit Installation
From the site "zircon.austin.ibm.com", logon using your DCE id/password,
and from the directory
"/:/projects/boss/dev_tools/java/linux/jdk118sr4", retrieve the file
"IBMJava118-SDK-1.1.8-4.0.i386.rpm". Use the "rpm" facility to install
the applications (i.e. "rpm -ivh IBMJava118-JRE-1.1.8-4.0.i386.rpm").
This will install the application in "/usr/bin", so if you wish to compile
from any directory you must add this directory to your "PATH" environment
variable in your ".profile" (or whatever script you use to set up your
environment)
WebSphere Application Server Installation on Linux
From the site "zircon.austin.ibm.com", logon using your DCE id/password,
and from the directory
"/:/projects/boss/botp/build/(your version)/linux/zipper"
(where the author installed "ptf2a0034.06" version) and retrieved the "IBMWebAS-3.02ADV_DOM-2.i386.rpm"
build. Using the "rpm" utility, install the build (i.e. "rpm -ivh IBMWebAS-3.02ADV_DOM-2.i386.rpm").
The environment variable "JAVA_HOME" must be set to point to the appropriate
java directory, either by exporting it from the command line or including
it in your ".profile" or other setup script (i.e. "export JAVA_HOME=/usr/jdk118"
(no quotes)).
Configuring Linux for Compiling and Building
Dependant upon the type of Linux install you have done, the C compiler and associated make and debug utilities might already be installed.
Otherwise, install the Red Hat Linux cd, and in the "Gnome RPM" window,
scope into "Packages", "Development", and "Languages", and select the following
icons (press ctrl to select multiple icons) "egcs 1", "egcs-c++ 1", "egcs-g77
1", and "egcs-objc 1", and then click the "Install" button on the top left
of this window.
WebSphere Linux Build Environment
From the site "zircon.austin.ibm.com", logon using your DCE id/password,
and from the directory
"/:/projects/boss/botp/build/(your version)/linux/zipper"
(where the author installed "ptf2a0034.06" version) and retrieve the "IBMWebAS-3.02ADV_DOM-2.i386.rpm"
and retrieve the appropriate zip file that contains the code you wish to
modify (i.e. "ncf.zip", "kansas.zip", and/or "prereqs.zip").
Now go to Building And Compiling.
WebSphere
Released Builds and Fixpacks
Note: this is an out of date list!
Version 302
PTF Version DateIBM Build Number
PTF1
4/14/00 ptf0015.03 (Adv) ptf0019.01
(Std)
PTF2
8/18/00 ptfa0034.05 (Linux =
34.06)
PTF3
1/8/01 ptf3a0051.02 (12/19
Build)
GM
7/31/00
GA
8/31/00
Version 3.5
PTF Version DateIBM Build Number
PTF1
9/7/00 ptfa0036.06
PTF2
10/16/00 ptf2b0041.07
ptf2b0042.03 (AIX)
PTF3
Mid Feb
This table contains the WebSphere Product Numbers, the associated
Release/version information, and the Product name/platform.
Component ID FESN
Releases End of Service Product
Name
5639F2800 0905647 100 110
01/31/01 Websphere AppServer
AIX
5639F2801 0905648 100 110
01/31/01 Websphere AppServer
NT
5639F2802 0905649 100 110
01/31/01 Websphere AppServer
SUN
5639F2803 0901329 100 110
01/31/01 Websphere AppServer
OS/2
5648B7900 0906173 200
01/31/01 Websphere AppServer
Advanced AIX
5648B7901 0906187 200
01/31/01 Websphere AppServer
Advanced NT
5648B7902 0906188 200
01/31/01 Websphere AppServer
Advanced SUN
5648B8200 0906172 200
01/31/01 Websphere AppServer
Standard AIX
5648B8201 0906185 200
01/31/01 Websphere AppServer
Standard NT
5648B8202 0906186 200
01/31/01 Websphere AppServer
Standard SUN
5648D0208 0906870 203
01/31/01 Websphere AppServer
Standard Linux
5648C8300 0906404 300, 350 12/31/01,
12/31/02 Websphere AppServer Standard AIX
5648C8301 0906405 300, 350 12/31/01,
12/31/02 Websphere AppServer Standard NT
5648C8302 0906406 300, 350 12/31/01,
12/31/02 Websphere AppServer Standard SUN
5648C8303 0906483 350
12/31/02 Websphere AppServer
Standard HPUX
5648C8307 0907339 300, 350
02/28/02,12/31/02 Websphere AppServer Standard LINUX
5648C8400 0906408 300, 350
12/31/01, 12/31/02 Websphere AppServer Advanced AIX
5648C8401 0906409 300, 350
02/31/01, 12/31/02 Websphere AppServer Advanced NT
5648C8402 0906410 300, 350
12/31/01, 12/31/02 Websphere AppServer Advanced SUN
5648C8403 0906485 350
12/31/02 Websphere AppServer
Advanced HPUX
5648C8404 0906486 300
05/31/01 Websphere AppServer
Advanced Novell
5648C8405 0906843 300
12/31/01 Websphere AppServer
Advanced Versata
5648C8407 0907338 300, 350
02/28/02,12/31/02 Websphere AppServer Advanced LINUX
5648C6501 0906553 300, 350
12/31/01, 12/31/02 Websphere Enterprise Edition AIX
5648C6500 0901446 300, 350
12/31/01, 12/31/02 Websphere Enterprise Edition NT
5648C6502 0906554 300, 350
12/31/01, 12/31/02 Websphere Enterprise Edition Solaris
5733WA200 0906620 302, 350
12/31/01, 12/31/02 Websphere AppServer Adv. AS/400 (56Bit)
5733WA300 0906643 302, 350
12/31/01, 12/31/02 Websphere AppServer Adv AS/400 (128Bit)
5697E3900 0906011 100
01/31/01 Websphere Studio WinNT
5697E3901 0906044 100
01/31/01 Websphere Studio Win95
5697E3902 0906045 100
01/31/01 Websphere Studio Win98
5697F6000 0906503 300
12/31/01 Websphere Studio WinNT
5697F6001 0906504 300
12/31/01 Websphere Studio Win95
5697F6002 0906505 300
12/31/01 Websphere Studio Win98
5697G3700 0907248 350
12/31/02 Websphere Studio Advanced
WinNT,95,98,2000
5697G3800 0907266 350
12/31/02 Websphere
Studio Pro WinNT,95,98,2000
5639I3300 0906465 300, 350
01/31/01, 02/28/02 Websphere Site Analyzer NT
5639I3301 0906466 300, 350
01/31/01, 02/28/02 Websphere Site Analyzer AIX
5639I3302 0906467 300, 350
01/31/01, 02/28/02 Websphere Site Analyzer SUN
5639I3303 0907250 350
02/28/02 Websphere
Site Analyzer WIN
5639I3307 0903152 350
02/28/02 Websphere
Site Analyzer LINUX
5765E4800 0907035 350
02/28/02 Websphere
Personalization AIX
5765E4801 0907036 350
02/28/02 Websphere
Personalization NT
5765E4802 0907037 350
02/28/02 Websphere
Personalization SUN
5765E4803 0907038 350
02/28/02 Websphere
Personalization HP-UX
5733A4700 0907159 350
02/28/02 Websphere
Personalization AS/400
5648B7800 0906012 133 136, 312
01/31/01, 02/28/02 IBM Http Server AIX
(312 = 1.3.12)
5648B7801 0905856 133 136, 312
01/31/01, 02/28/02 IBM Http Server SUN (312
= 1.3.12)
5648B7802 0906192 133 136, 312
01/31/01, 02/28/02 IBM Http Server NT
(312 = 1.3.12)
5639D54L0 0900627 462
01/31/01 Domino
Go Webserver Win95
5697A37L0 0901435 462
01/31/01 Domino
Go Webserver OS/2
5697A43L0 0901615 462
01/31/01 Domino
Go Webserver WinNT
5697A45L0 0901617 462
01/31/01 Domino
Go Webserver SUN
5697A47L0 0901618 462
01/31/01 Domino
Go Webserver HP-UX
5697A74L0 0901619 462
01/31/01 Domino
Go Webserver AIX
Installing
AIX from Scratch
Dependant upon the type of machine you are installing on, the bootup
sequence could be different. Regardless of which type, insert the AIX CD
1 in the machine ( This CD must be in the machine during the bootup process),
and initiate the boot (the LCD display should reflect "OK", press the white
button on the front of the box)
On later machines, press "F1", click on the "Multiboot" icon , then click on "Install From" icon, select "CD", then select "Install".
(OR)
As soon as the icons begin to display on bootup, press the F5 key. After a few minutes you will get a text screen, press F1 key and press "Enter". At the next screen accept the default by pressing enter ("1 Type 1 and press Enter to have English during install.")
You will get a menu:
1) Default Settings ....
2) Change/Show ......
3) Start .....
Select option 2, Change/Show ....., and get the Change Method Of Installation menu. Select 1 (Complete Overwrite). If you have more than one hard disk drive, you should select 2 to do both disks. Press "Enter" and you will return the "Installation and Settings" screen. The default choice "[0]" should be selected, so press "Enter". Wait a month while software is installed. After the software is installed the system will reboot itself, and then you will get a "Configuration Assistant TaskGuide" window.
This includes the message 'NOTE: Your System's paging space is less than the recommended amount. Insufficient paging space can cause serious operational problems. To increase paging space, select "Manage system storage and paging space" from the list on the next panel.'
The GUI window "Configuration Assistant TaskGuide" will present several tasks which should be performed (at this point you must learn to squint!):
Set or verify system date and time
Set password for administrator (root user)
Manage system storage and paging space
Configure network connection
Update installed software after a migration installation
Configure Web-based System Manager to run in a web browser (applet
mode)
Configure Online Documentation and Search
Exit the Configuration Assistant
Press "Next" and in the set date window, set the date. Press "Next"
and go to previous screen.
Press "Next" and in the set password window, set root password. Press
"Next" twice to get back to
"Configuration Assistant TaskGuide" window.
Press "Next" and in the manage system window, select default recommended
space by pressing "Next"
(384 or 1024 Mbytes)
Press "Next" and in the configure network window, select "Manually
configure TCP/IP" and press "Next"
Enter the host name, IP address, and subnet mask returned from
the above web site request and
press "Next"
A window "Configure TCP/IP TaskGuide" will be displayed regarding multiple network adapters, assume you have one and select the "Token Ring Network Interface". In the next screen "Ring Speed", use "autosense" and press "Next".
The next window is used to specify the default gateway address. Use the values from the IP address requested previously, i.e.:
Default gateway address: 9.27.224.1
Domain Name: raleigh.ibm.com
IP address of name server: 9.37.0.6
You will then get a confirmation screen about the above entered IP information, press "Next". A popup window will be displayed showing progress, when "Finished" appears in the menu bar of this popup, press "Close". In configure TCP/IP taskguide window, press "Next".
The next selection is the "Update installed software" section, which will not be required as you have just installed a new version.
If you wish to do systems administration using a web based System Manager, select it and "Next". (yer on your own on this one. The next set of screens require a server name, and since this is a new install, there probably aint one to select).
The next task selected is the "Configure Online Documentation and Search". This will automatically install a Netscape browser which can be used for viewing documentation on the AIX box, but if you want the documentation to be viewable from other machines you must install a web server.
The following assumes you want to do this install.
On the "Configure Online Documentation and Search window, click "Next". In the next screen select "Setup the Documentation Library Services on this computer", and click "Next". In the next screen select "Yes, remote users need to use the Documentation Library Services from this computer" and click "Next". In the next screen select "Yes, use Netscape" and click "Next". In the next screen select "Install the Netscape Browser" and click "Next". In the next screen you will be prompted to put the "4.3 Bonus Pack CD" into the CD drive. Do this, then select the physical device "/dev/cd0" and click "Nex"". A "Working" pop-up window will be displayed showing progress. When finished, click "Close".
Select "Exit the Configuration Assistant" and press "Next", use the default "Finish now, and do not start Configuration Assistant when restarting AIX." and press "Finish".
Dependant upon the type of AIX installation process you followed, this could have been started using a Netscape browser interface. If so, click "exit" and "exit from the welcome" center" hot link to end. This will put you back into the Configuration Assistant. Select "Exit the Configuration Assistant" and click "Next".
You will then be prompted for you user id and password (root/yourpassword)
Note: you can reopen Configuration Assistant by typing "configassist"
on the command line.
Instructions to Setup Mainframe Access
This section describes how to setup the Windows NT terminal emulator
to connect to a mainframe. You must have an account, Id, and password for
use of this emulation, and have installed the "IBM Personal Communications"
software package.
Mainframe
Access Software Download
Creating
a Session Macro
Mainframe Access Software Download
To load the Personal Communications Software from the IBM Standard Software
installer web page connect to url http://w3-1.ibm.com/download/standardsoftware/index.html
and select a location (North America). In the bar at the top of this page
select "Windows NT". In the "WINDOWS NT PRODUCT CATALOG" web page
select "All Applications", then select "IBM Personal Communications v4.3.1.1".
Select the drive you wish the software to be installed on and click "Install
No"". In the next pop-up, select "Save to disk", and use the default
path. After the installation you will probably have to restart your computer.
1) From the desktop, with the right mouse button, click on [Start],
slide to [Programs], then
[IBM Personal Communications], and select "Start or Configure
Session" (read pop-up window
and click [OK]).
2) In the pop-up window "Customize Communication", in the "Host:" section,
select [S/390]
then click on the [TCP/IP] icon. A "Telnet3270" window will appear,
where you will type
the name "sersna" in the "Host Name or IP Address" field and
then click the [OK] button.
3) Click [OK] in the "Customize Communication - 3270" window, [OK] in
the "Customize Communication"
window and you will be connected to the "IBM/Global Services - Southeast
Region". (Note: If you have problems connecting to "sersna1", try [Communications]
[Disconnect] [Configure] and try "sersna2".)
4) Enter the vm node you wish to connect to (i.e. "RALVMS"), and press
"enter" ("Ctrl" on most PC keyboards). Enter your ID and password, and
"enter"
(At this point you have the option of saving this connection as an icon. To do so, from the top menu bar select "File", pull down to "Save As", and enter the name you wish for your icon. You will be queried whether you want to add an icon for this session to a folder. If you click [yes], another pop-up window will be displayed where you can add the icon to a folder, or "check" the "Add icon into the folder of Desktop" which will create an icon on your desktop when you click the [OK] button.)
Note: If you have problems connecting to "sersna1", try [Communications]
[Disconnect] [Configure] and try "sersna2".
You can create macros to execute the normal sequence of commands for a particular mainframe session. This is done by using the "Record" mechanism within the Personal Communication System (PCS).
Double click on the icon you created in the previous section, or from a previously created session click the "Record" button at the top of the session window. You will be prompted for a filename for your macro. Once you have selected a filename you are in the record mode. Go through the sequence of commands you wish to record (i.e.go through the set of Retain commands required to get you to the "read" section of a session), then click the "Stop" button at the top of your PCS session window.
From then on you should be able to simply double click on your session
icon, wait for the window to popup, and doubleclick on the "Play" button
on the top of your PCS session window. You will be queried for the macro
name, select it and clidk "OK". Your macro should begin executing all the
commands you recorded.
Instructions to Setup PC to Workstation Access
This section describes how to setup the Windows NT terminal emulator to connect to a workstation. You must have an account, Id, and password for use of this emulation, and have installed the "Exceed" software package. Once you have setup Exceed, you will be able to logon to a Unix/AIX/Linux machine from your NT desktop. Once logged on to the remote machine, you can invoke the WebSphere Administration Console GUI and have it displayed on your Windows machine and perform normal administrative activity.
All of the following instructions assume you install Exceed using defaults.
Instructions for Installation and Configuration of Hummingbird Exceed Software
1. Install CD, go to E:Winnt/Exceed and run "Setup" and follow setup instructions.
(or)
From Windows Explorer, in the menu bar
select "Tools", and "Map Network Drive"
(Normally a drive will be preselected)).
In the "Path" field, enter "\\9.67.133.217\software"
(no quotes), and in the "Connect as:"
field enter "id4soft" (no quotes), (then click "OK").
At this point (if the connection was
successful), this drive will then appear in your Windows
Explorer window. When you go to this
machine to get software you will be queried for a
password "licenseme" (no quotes). Scope
down to:
"non-Ibm/Exceed_6.2_from_Hummingbird/EXCEED/NT_INTEL/SOFTWARE"
and double click on "Setup" and follow
the setup instructions.
2. From the NT "Start" button, select "Programs", "Exceed",
and "Xconfig". The "excdeed.cfg - Xconfig"
window will pop up, double click on
the "Screen Definition" icon. In this window, within the "Screen0"
tab,
under the "Window Mode", select "Multiple",
and in the "Window Manager" field, select "Native", then
click "OK"
3. From the NT "Start" button, select "Programs", "Exceed",
and select "Xstart". In the "Untitled - Xstart
Client Startup Application" window,
enter in the following fields to appropriate data:
Start Method:
REXEC (TCP/IP)
Program Type: X Window
Login/Userid: (your
account name on the Unix/Linux/AIX server)
Login/Password: (your password)
Login/Host:
greenk.charlotte.ibm.com (this is domain name for the server)
(in the following field, substitute YOUR PC’s IP address, and
do not forget the ":0.0".
The title is optional, and if you wish to use it, substitute "yourtitle"
for what you wish)
Login/Command: @(XTerm, method=stdappdb) -title yourtitle
-display 9.27.62.179:0.0 -sb &
Login/Prompt: (click "none"
radio button (= on))
4. From Xstart menu bar click "File" and "Save As" and create a filename for your configuration.
5. From Xstart menu bar click "Run" to start an Xwindow.
6. To add program group from Xstart menu bar click "Install",
and Program Group Description
should be equal to filename created
above. Click "OK". Program Group window should
appear. Click on filename/icon to invoke
creation of a window.
Note: When logging onto some AIX machines your ".profile" might not be executed. If this is the case, it might be necessary to issue the command:
. ./.profile (that's dot, one space, dot, forward slash, dot, "profile")
If there is a "export DISPLAY" assignment in this profile, it might
change the IP address which you might have to reassign to your desktop
machine (i.e. "export DISPLAY=9.27.62.179:0.0" (no quotes, use your desktop
machine's IP)).
For cable modem (i.e. "Roadrunner"), you must install the "Sign On"
software, which can you can get from IBM ISSI or you can order the ISSI
CDs, but you need to register to SINE (which is the "Sign on" software
for Road Runner). You will use your normal IBM Global Network ID
and password for this access.
Installing DB2 Universal Database on AIX
This procedure assumes you have a CD with DB2 on it (i.e. DB2 Universal Database Personal Edition Versino 7.1). This could be a CD with only DB2 on it, or the Advanced WebSphere Installation Disk. To access either of these from your AIX machine you must first mount it (see instructions "Mounting a CD on AIX"). For DB2 CD, go to the section "DB2 Universal DataBase CD Installation", then "DB2 Installation Instructions". For WebSphere CD DB2 installation go to "Advanced WebSphere CD DB2 Installation", then "DB2 Installation Instructions". For either method you must mount the appropriate CD so AIX can access it.
Different versions of WebSphere require different prerequisite version
of DB2 and associated fixpacks. The following examples describe both cases,
but it is up to you to assure you have the correct versions.
Advanced
WebSphere CD DB2 Installation
DB2
Universal DataBase CD Installation
DB2
Installation Instructions
Installing
DB2 Fixpacks
Problems
With Installation/Configuration of DB2
You will also need a large amount of disk space for DB2, a minimum "/usr"
file system size of 3809280, "/home" to a size of 81920, and "/" to a minimum
of 65536.
Miscellaneous anomalies (hopefully you can ignore this!): After
invoking the script to install db2, the installer complained about not
having the 2 environment variables listed below set. Also, attempts were
made to install DB2 which were unsuccessful due to not being able to modify
"/etc/services". My final attempt was done without the environment variables
as well as no modification to "/etc/services".
If so, you need to first remember the file permissions of, then chmod
"etc/services" from "-rw-r—r—" to "777". After the install, change it back
to the original permissions.
Advanced WebSphere CD DB2 Installation
Before installing DB2 you must first set 2 environment variables at the command line:
export DB2IRCX=/CD/aix/DB2/install/locale/C/db2inst.rcx
export DB2ICAT=/CD/aix/DB2/install/locale/C/db2inst.cat
After mounting the CD, go to the directory "/CD/aix/DB2/install" and at the command line, execute the command "./db2inst". You should be presented with a screen "Install DB2 V5". Using the "Tab" key, go to the "[ ] DB2 Client Application Enabler" field, and press "Enter" to select the installation of this component. Also select the "[ ] DB2 UDB Enterprise Edition" and "[ ]DB2 Connect Enterprise Edition". Using the arrow keys go to the "[ OK ]" field and press "Enter"
Now execute the instructions in the following section "DB2 Installation
Instructions".
DB2
Universal DataBase CD Installation
Before installing DB2 you might need to first set 2 environment variables at the command line (you will be told to do this if necessary by the invocation of the db2setup script):
export DB2IRCX=/CD/db2/install/locale/C/db2inst.rcx
export DB2ICAT=/CD/db2/install/locale/C/db2inst.cat
From the directory "/CD", from the command line execute the command "./db2setup". You should be presented with a screen "Install DB2 V6.1". Using the "Tab" key, go to the "[ ] DB2 Administration Client" field, and press "Enter" to select the installation of this component. Also select the "[ ] DB2 UDB Enterprise Edition", "[ ]DB2 Connect Enterprise Edition", and "[ ] DB2 Software Developer's Kit". Using the arrow keys go to the "[ OK ]" field and press "Enter"
Now execute the instructions in the following section "DB2 Installation
Instructions".
In the "Create DB2 Services" screen, using the "Tab" key, go to the "( ) Create a DB2 Instance" field and press "Enter".
In the "Create DB2 Services" "DB2 Instance" screen, select "[ OK ]" to select the default settings. (If you do not input a password, a pop-up screen will appear stating a system-generated one will be used (ibmdb2).) Click "[ OK ]" in the pop-up and "[ OK ]" in for the "Create a DB2 Instance" screen.
At this point you might be prompted to specify "User-Defined Functions",
or "Fenced User" so use the defaults, tab to "[ OK ]" and press
"Enter". (If you do not input a password, a pop-up screen will appear
stating a system-generated one will be used (ibmdb2).) Click "[ OK
]" in the pop-up and "[ OK ]" in for the "Create a DB2 Instance"
screen.
Back in the "Create DB2 Services" screen, tab to the "( ) Create
the Administration Server." field and press "Enter". In the "Administration
Server" screen, select "[ OK ]" to select the default settings.
(If you do not input a password, a pop-up screen will appear stating a
system-generated one will be used (ibmdb2).) Click
"[ OK ]" in the pop-up. Another pop-up will appear stating
"DB2SYSTEM will be set to '(yourservername)'. Click "[ OK
]" and you will return to the "Create DB2 Services" screen. "Tab to the
"[ OK ]" field and press "Enter". A "Summary Report" will
be displayed showing product components to be installed. Click "[ Continue
]". A "last chance" pop-up will appear, select "[ OK ]". Wait
a month.
For DB2 6.1, when this is complete you will need to install DB2 version 6 FixPack 4 which resides on "id10t.raleigh.ibm.com" in directory "/home/aixsw/db2.6.fixpak4" (use logon id/password guest/guest).
For DB2 5.x, install some other fixpack ( I don’t know which one).
These instructions describe installing fixpack 4 for DB2 6.1. They should be similar for 5.x and it’s associated fixpacks.
Before the install you must first stop each instance of DB2 as well as the DB2 Administration Server. First log on as the instance (i.e. "su - db2inst1"), and do the following:
. $HOME/sqllib/db2profile
db2 force applications all
db2 terminate
db2stop
db2licd end (run at each physical node)
exit
and for the admin instance, log on as db2as (i.e. "su - db2as") and type the following:
. $HOME/sqllib/db2profile
db2admin stop
exit
/usr/sbin/slibclean
Using ftp, copy the files from "id10t.raleigh.ibm.com" in directory "/home/aixsw/db2.6.fixpak4" to a working directory on your machine (i.e. "/tmp/db2.fixpak4"). From the command line type "smit update_all", and when prompted, enter the name of the directory the fixpack files reside in, and click "OK". In the next window, using default values, click "OK". You will get an "ARE YOU SURE" pop-up, click "OK". The software will be installed, with the status of the install being displayed in a popup window. When the update is done, click "Done" and you will be returned to the Update Installed Software window, click "Cancel" to leave this window.
As root, from the command prompt issue the command "/usr/lpp/db2_06_01/instance/db2iupdt db2inst1", which will run a utility to update the instance. Wait several seconds and you should get a "completed successfully" message.
Then update das using the command "/usr/lpp/db2_06_01/instance/dasiupdt db2as". Wait several seconds and you should get a "completed successfully" mess"ge.
Login as instance owner (i.e. "db2inst1") and issue the command "db2start". Wait several seconds and you should get a "processing was successful" message. Also as instance owner or db2as user, start admin by issuing the command "db2admin start". You should get a "started successfully".
As root, you will then need to use smitty or smit to add 3 groups to your root account, "db2asgrp", "staff", and "db2iadm1", as well as add "/home/db2inst1/sqllib" to your root ".profile" "PATH".
Issue the command "db2 create database was" from the command prompt
using the root account to create your Websphere database. You must
the change the "heap size" of this database by issuing the command
"db2 update db config for WAS using applheapsz 256". (you might need
to change this to 512)
While logged in as root, switch user to "db2as" (i.e. "su - db2as", or your db2 admin server user id), and from the command prompt type "db2admin start" (and then wait several months!). After this command is successfully executed (the command line prompt will return), type "exit" to return to the root account. Switch user to "db2inst1" (i.e. "su - db2inst1", or your db2 instance user id), and from the command prompt type "db2start".
(?1) as root, /usr/lpp/db2_06_01/instance
Problems With Installation/Configuration of DB2
After dropping the database, if you try to delete the database it, it
says it doesn’t exist, but if you try to create it, it says it already
exists. To cure this type "uncatalog database (database name)".
Logon as root, look for ps –ef |grep db2 kill –9 start with db2sysc
(main process) "ipcs |grep db2", "ipcrm –(mq or s) then the id of
the process (i.e. ipcrm –m 3). Update the instances as root
cd /usr/lpp/db2_05_00
And go into instance directory. In there a utility "db2iupdt", run
it against "db2iupdt db2inst1".
Then update das using "dasiuptd db2as" update das instance.
Login as instance owner and type "db2start". And also as instance owner
or db2as user start admin by "db2admin start"
Installing DB2 Universal Database on NT
This procedure assumes you have a CD with DB2 on it (i.e. DB2 Universal Database Personal Edition Version 7.1). This could be a CD with only DB2 on it, or the Advanced WebSphere Installation Disk (obtain from the Test Lab). To access either of these from your AIX machine you must first mount it (see instructions "Mounting a CD on AIX"). For DB2 CD, go to the section "DB2 Universal DataBase CD Installation", then "DB2 Installation Instructions". For WebSphere CD DB2 installation go to "Advanced WebSphere CD DB2 Installation", then "DB2 Installation Instructions". For either method you must mount the appropriate CD so AIX can access it.
Different versions of WebSphere require different prerequisite version
of DB2 and associated fixpacks. The following examples describe both cases,
but it is up to you to assure you have the correct versions.
DB2 Samples Database Creation on NT
From the NT "Start" button, select "Programs", "DB2 for Windows NT", and click on "First Steps". This will pop-up a window in which you should first click on the "Create the Sample Database", the creation of which might take several minutes.
After the pop-up appears stating the operation is complete, you can
perform 4 different actions, "Create the Sample Database", "View
the Sample Database", "Work with the Sample Database", or "View the Product
Information Database". View the database by clicking on the "View the Sample
Database" icon. You will be queried for the DB2 id and password.
A "Command Center" window will be displayed which has the following text,
which includes instructions on executing the script, followed by the results
of executing the script:
-- This DB2 command script lists each employee in the SAMPLE database.
-- You can use it to confirm that SAMPLE was created successfully.
--
-- To run the script, select "Script" from the main menu of the
-- Command Center, then select "Execute". You can also use the
-- "Gears" push button at the top left of the command center.
-- The list of employees will appear in the "Results" page.
-- To make changes to the script, simply return to the "Script" page.
connect reset;
connect to sample;
select firstnme, lastname, empno from employee order by firstnme;
**** RESULTS OF EXECUTING THE SCRIPT ****
;
---------------------------------------------------------------------------------------------------------
SQL1024N A database connection does not exist. SQLSTATE=0800
Database Connection Information
Database server = DB2/NT
6.1.0
SQL authorization ID = SHAYDEN
Local database alias = SAMPLE
FIRSTNME LASTNAME EMPNO
------------ --------------- ------
BRUCE ADAMSON
000150
CHRISTINE HAAS
000010
DANIEL SMITH
000250
DAVID BROWN
000200
DOLORES QUINTANA
000130
EILEEN HENDERSON
000090
ELIZABETH PIANKA
000160
ETHEL SCHNEIDER
000280
EVA PULASKI
000070
HEATHER NICHOLLS
000140
IRVING STERN
000060
JAMES WALKER
000190
JAMES JEFFERSON
000230
JASON GOUNOT
000340
JENNIFER LUTZ
000220
JOHN GEYER
000050
JOHN PARKER
000290
MARIA PEREZ
000270
MARILYN SCOUTTEN
000180
MASATOSHI YOSHIMURA
000170
MAUDE SETRIGHT
000310
MICHAEL THOMPSON
000020
PHILIP SMITH
000300
RAMLAL MEHTA
000320
SALLY KWAN
000030
SALVATORE MARINO
000240
SEAN O'CONNELL
000120
SYBIL JOHNSON
000260
THEODORE SPENSER
000100
VINCENZO LUCCHESSI
000110
WILLIAM JONES
000210
WING LEE
000330
32 record(s) selected.
Netscape Enterprise Server and IPlanet Installation
Netscape will need a group and userid created for it's operation (see
section "Using The Systems Management Tool to Add Users and Groups" for
instructions on creating these). (Note: the group and id should reflect
the Netscape version (i.e. "es36" for version 3.6 or "es40" for version
4.0).
Netscape
3.63. Installation
Iplanet
Installation
Creating
a Netscape Server Instance
Netscape
3.6 IPlanet Management Server Instance Creation
Netscape
4.0 IPlanet Management Server
Instance
Creation
For this installation you will need the Netscape Enterprise Server CD or tar file, the tar file ("enterprise-3.63-domestic-us.powerpc-ibm-aix4.1.5.0.tar", or "ent40sp2daix.tar" for Netscape 4.0/Iplanet) is available on "id10t.raleigh.ibm.com" in the "/home/aixsw/ns.363" or "ns.40" directory .
Begin by using ftp to copy the file to your machine, and make/put it in the directory you wish to install in (i.e. "/usr/ns363"). The run tar on the file (i.e. "tar -xvof enterprise-3.63-domestic-us.powerpc-ibm-aix4.1.5.0.tar"), which will put all of the files in "enterprise-3.63-domestic-us" (I renamed this directory "es363" due to intense laziness). Change to this directory and run the setup script (i.e. "./ns-setup")
Agree to the license terms ("yes") and press "Enter", either use the default directory or change it (I use "/usr/ns363/es363") and press enter. You then will be queried for your domain name. Enter your full machine name "idiot.raleigh.ibm.com" and continue by pressing "Enter".
You will the be queried for the userid and group created previously ("es363" and "es363"). You will then be queried regarding LDAP directory services. Use the default [n]. You will then be queried regarding usage of a local database rather than LDAP. Press continue (then wait a month).
You will then be queried for an "Administration port", with a default provided. Then you will be queried for whether you wish to run your administration server as root, select the default ("[root]"). You will then be queried for the path of a 2.x version of Netscape Administration Server, choose the default ([/usr/ns-home]"].
You will then be queried for the Administrator ID, choose the default ("[admin]"). The next field is the password for the Administrator. I recommend you use the same password as the root password on your machine. You will be required to enter this password twice. Press "enter" and the installer will perform some actions and require you to press "Enter" to continue.
You will then be queried for a browser. As you will probably be using your NT browser, accept the default or ignore this.
At this point your Netscape Administration Server has been created and
started. You should be able to go to any browser in the Raleigh domain
and connect to your Admin Server by specifying your machine and admin port
number (i.e. "id10t.raleigh.ibm.com:6969"). You will be queried for the
user and password (i.e. "admin" and your password).
For this installation you will need the Netscape IPlanet Server CD or tar file, the tar file ("ent40sp2daix.tar" for Netscape 4.0/Iplanet) is available on "id10t.raleigh.ibm.com" in the "/home/aixsw/ns.4.0" directory .
Begin by using ftp to copy the file to your machine, and make/put it in the directory you wish to install in (i.e. "/usr/ns40"). The run tar on the file (i.e. "tar -xvof ent40sp2daix.tar"), which will put all of the files in "ent40sp2daix " (I renamed this directory "es40" due to intense laziness). Change to this directory and run the setup script (i.e. "./setup"). You will get a screen of installation info/recommendations and be queried if you wish to continue, press "Enter" to continue.
Agree to the license terms ("yes") and press "Enter". You will then be queried for the type of installation, choose "2" Typical installation". You will then be queried for where you want the installation to reside, either use the default directory or change it (I use "/usr/ns40/es40") and press enter. You will then be queried for the components you wish to install, select all (default). In the next screen, select all components again (default).
You then will be queried for your domain name. Enter your full machine name "idiot.raleigh.ibm.com" and continue by pressing "Enter".
You will the be queried for the userid and group created previously ("es363" and "es363"). You will then be queried as to how to run the Administration server, choose the default (root).
You will then be queried for the Administrator ID, choose the default ("[admin]"). The next field is the password for the Administrator. I recommend you use the same password as the root password on your machine. You will be required to enter this password twice.
You will then be queried for an "Administration port", with a default provided (I used "6969").
You will then be queried for an "Enterprise Server port", with a default
provided (I used "714")
You will then be queried about registering with an existing Directory
Server, choose "No" (default). You will then be queried about the Enterprise
Server’s Content Root, select the default.
You will then be queried about a JDK installation. As WebSphere will use it’s own version, select to use your own ("yes"). You will then be asked for the location of your JDK, and as you have not installed one yet, click "Enter" (a JDK will be installed anyway).
You will then be queried about the installation cache "install.inf", it’s up to you on this one.
You will get a message "Go to /usr/ns40/es40 (or your directory name)
and type "startconsole" to begin managing your servers. Invocation of this
will cause your Netscape browser to come up and you will be queried for
the user and password (i.e. "admin" and your password).
Creating a Netscape Server Instance
You must first create a user for your server instance (use the above instructions for creating a user), but supply the "Primary GROUP" as "staff" and "Group Set" as "staff" and the Netscape group for your version of Netscape (i.e. "es363" for Netscape version 3.63, or "es40" for Iplanet). Make the "HOME director" the same name as your webserver server id (i.e. "/home/level3").
Netscape 3.6x Managment Server Instance Creation
From your Admin Server browser session, select "Create New Netscape Enterprise Server. Enter your server name (i.e. "level3"), a "Server Port" (i.e. "10001"), the "Server Identifier" (i.e. "level3"), and the server user (i.e. "level3"). Use the default radio button selections (only "Never attempt to resolve IP addresses into host names" being selected), and change the "Document Root" to include the server instance directory name (i.e. "/usr/ns363/es363/https-level3/docs"). Then click "OK
If you select "Return to Server Administration", you can see your server and start it by clicking the "Off/On" switch. Dependant upon whether you have manually modified your Netscape server instance's configuration files (which WebSphere does upon installation with the associated plugin), this might not work. If it does not, you can go to your server directory (i.e. "/usr/ns36/es36/https-yourserverinstancename/") and execute the "start" script (i.e "./start" (no quotes)).
You may then use a browser and connect to your server instance (i.e.
"http://id10t.raleigh.ibm.com:1001"). You should change the file mode and
group to your userid and "staff" respectively, while logged on as root
(i.e. "chgrp -R staff https-level3" and "chown -R level3 https-level3"),
so this user can modify the files within their directory as required.
Netscape 4.0 IPlanet Management Server Instance Creation
From your Admin Server browser session, in the "Manager Servers" browser window section, click "Add Server". Enter your server name (i.e. "level3"), a "Server Port" (i.e. "10001"), the "Server Identifier" (i.e. "level3"), and the server user (i.e. "level3"). Use the default radio button selections (only "Never attempt to resolve IP addresses into host names" being selected), and change the "Document Root" to include the server instance directory name (i.e. "/usr/ns363/es363/https-level3/docs"). Then click "OK" and your server instance will be created.
Click "Manage Servers", and in the "Select A Serve"" selection, select your server and click "Manage". In the server startup window click "Server On" to start your server. Dependant upon whether you have manually modified your Netscape server instance's configuration files (which WebSphere does upon installation with the associated plugin), this might not work. If it does not, you can go to your server directory (i.e. "/usr/ns40/es40/https-yourserverinstancename/") and execute the "start" script (i.e "./start" (no quotes)).
You should change the file mode and group to your userid and "staff"
respectively, while logged on as root (i.e. "chgrp -R staff https-level3"
and "chown -R level3 https-level3"), so this user can modify the files
within their directory as required.
Get the tar file using ftp from id10t.Raleigh.ibm.com using guest/guest.
The tar file is in the directory
/home/aixsw/ns.comm..4.7 filename "communicator-v47-us_ppc_rs6000-ibm-aix4.tar".
Untar this and go to the resulting directory created, and execute the command
"./ns-install". You will be queried for a directory to install to, use
the default. You will be asked if you wish it created, "y".
First begin by creating and changing to a temporary directory and ftp'ing to "id10t.raleigh.ibm.com" as user id/password "guest"/"guest". Go to directory "/home/aixsw/cc.3.6.6/tar.main" and retrieve the file "CCpp.366.tar". Untar the file and then use the "smitty" utility for the installation of this software. (The file can also be downloaded from "http://w3.torolab.ibm.com/~aixcpp/index.html")
Invoke "smitty", and go through the following sequence of panels:
Systems Management
Software Installation and Maintenance
Install and Update Software
Install and Update from LATEST Available Software
(specify the directory you untarred the cc.3.6.6 files)
(The smitty command was "installp –acgNQqwX –d /home/aixsw/cc.3.6.6/tar.main –f file 2&1")
This installation resulted in 3 filesets not passing requisites, memdbg.aix41.adt, ibmcxx.html.ja_JP.search, ibmcsxx.html.ja_JP.doc,
xlC.msg.en_US.rte was already supersceded by 4.0.2.0
Now installed the license using lum (see /license.stuff on id10t)
The following is crippled notes on installing C license, I will update/clean
when I have to do this again Steve
-- Installed and Configured License Use Management (LUM) V4
- LUM needs to be installed to allow
C for AIX V4
to work. LUM comes with
C FOR AIX V4, but does not include
the GUI Configuration tool.
You will have to either install
ipfx package OR use the
shell scripts provided.
- LUM is installed at /var/ifor.
- Follow the instructions for configuring
LUM V4
contained in the /usr/vac/README.password
file.
- Use the /var/ifor/i4config to configure LUM.
by typing ./i4config at the command line. The first thing you are
prompted for is the configuration. I said 4 both NEtwork and Nodelock
license server. Was queried for y/n on both, and said y for both.
- Issued ./i4blt to start GIU LUM and
enrolled
cforaix_cn.lic
- The above mentioned file does not include
the commands
to enroll the C for AIX
V4 product from the command line,
so it is:
i4blt
-a -n mortonm -f /usr/vac/cforaix_nc.lic
-R "'Michael Morton'" -T 1
i4blt -a
-n mortonm -f /usr/vac/c44aix_c.lic
-R "'Michael Morton'" -T 1
Once the C compiler is installed, you can compile a C source file by
entering the "cc" command followed by the input source and output
files (i.e. "cc SocketConnect.c -o SocketConnect" (no quotes)).
C/C++ Installation and Compiling/Building on NT
(This section describes how to setup Microsoft Test and Development (MTD) environment on NT. You might wish to use a different compiler. It also assumes you already have a JDK installed (see "JDK 1.1.8 Installation")
Obtain the Microsoft Test and Development Environment CD, and upon installation of the CD you might be queried to increase your registry size. If so, you should probably let it.
You will be queried for whether you accept the license, yes you do. Select the "accept" button and click "Next". In the next screen enter your name and "IBM" and click "Next". The next screen will query you about updating your JDK. You must let it to continue, select "Next". You must then reboot your machine.
After rebooting, you will be returned to the MTD Installation process.
Choose a folder or accept the default and then click "Next". The "Setup"
screen will recommend you stop all other applications (whatever.....).
Click "Continue", and then click "OK" in the product id window. For
the Custom installation, accept the default options, and click "Continue".
In the "Setup Environment Variables", DO NOT select the "Register Environment Variables" option, selct "OK". In the "Visual SourceSafe Database Format" pop-up, make up your own mind, and select the appropriate option (the author said "Yes"). Wait a month.
You will then get a pop-up "Setup has installed an icon in the Mirosoft Visual C++ 6.0 Tools program group, that will allow you to install a subset of the Windows NT sysmte symbols (.DBG) files from your Visual C++ CD-ROM. If the symbols are not installed, the Visual C++ debugger is not always able to determine the context of called functions. For easier application debugging, it is strongly recommended that you install these files. See the Visual C++ Readme for more information."
You will then be queried for "Do you want to overwrite JIT settings", the author selected "NO". You must then restart windows, click "Restart Windows". Upon restart a new pop-up "Installation Wizard for Visual Studio 6.0 Enterprise Edition" will be displayed (the author did not have the CD to do this but went through the exercise anyway). You will be queried for whether you wish to install the documentation. The author selected no and then clicked "Next". You will then be queried for which client tools to install. Select " InstallShield" and click "Next". Int the "Server Setups" section, click "Next". You can then elect to register your software.
You can obtain the source files for a build from the Austin ftp site (zircon.austin.ibm.com, using your DCE/DAAT id/password, in directory /:/projects/boss/build/botp/(the version you wants name, i.e. "ASV302")/test/q00(yourversion).(your O.S)/(your O.S.)/zipper (the NT version of the build does not have an O.S. in the directory name). In the zipper directory resides the standard and advanced versions of build, US and domestic. This includes the normal "GUI" install (i.e. "ibmwebasv3_dom_adv.exe"). The source files for your version of build is in the "ncf.zip", "kansas.zip", and "prereqs.zip" files in this same directory.
The following is a NT batch file for setting up your environment for
building. You must modify this file to reflect your paths for your environment.
If you choose not to use the following file, you must manually perform
the actions within the file in the DOS window where you will execute the
build process (or alternatively set the parameters/values in your "System"
settings). Note that the third line in this batch file (the "set PATH"
statement) is actually on one continuous line.
Also note that this is for a WebSphere 3.023 version build so you might
need to modify it for other versions to reflect different *.jar files.
REM # Build environment setup batch file for NT
echo off
cd c:\ws3023.sources\ncf
set PATH=%PATH%;c:\ws3023.sources\ncf;c:\jdk1.1.8;c:\Program Files\Microsoft Visual Studio\Common\MSDev98\Bin
REM # For NT, you must have the appropriate version os WebSphere installed
and
REM # make the following assignments to the CLASSPATH environment variable
REM # This section adds all WebSphere *.jar files to CLASSPATH for
Java compilation..
set JARS_AT=c:\WebSphere\AppServer\lib
set TEMP_CLASSPATH=%JARS_AT%\admin.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\BEXruntime.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\admin.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\chart.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\console.jar;
set TEMP_CLASSPATH= %TEMP_CLASSPATH%%JARS_AT%\databeans.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\deployTool.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\dertrjrt.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\developBEX.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\developEJB.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\ejs.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\ejscp.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\ibmjndi.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\ibmwebas.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\iioprt.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\iioptools.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\ivbjfaceall.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\jacl.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\jetace.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\jskit-cfwk.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\jsp10.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\lotusxsl.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\ns.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\repository.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\servlet.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\servletconsole.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\sslight.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\swingall.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\tasks.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\tcljava.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\ujc.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\vaprt.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\webtlsrn.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\x509v1.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\xml4j.jar;
set TEMP_CLASSPATH=%TEMP_CLASSPATH%%JARS_AT%\xml4j_1_1_16.jar;
set CLASSPATH=%TEMP_CLASSPATH%%CLASSPATH%
set JANUSPATH=c:\ws3023.sources\ncf\ncf
set JAVA_HOME=c:\jdk1.1.8
cd c:\Program Files\Microsoft Visual Studio\VC98\Bin
vcvars32.bat
# The following line is optional
cd c:\ws3023.sources\ncf
After executing the above batch file, you can perform WebSphere NT builds.
You can do a full build by going to the WebSphere "ncf" directory or alternatively
a partial build by going to the appropriate subdirectory and executing
the "makent cleanall" followed by a "makent all" commands (if the cleanall
does not work and you must manually delete all *.dll files before executing
the "makent all" command).
Compiling DLL's
In some directories the actual compiled code (*.dll) will be in a directory
UNDER the directory you compiled within (i.e. "Release" directory). Once
you have compiled your code you must stop the WebSphere Admin server as
well as the web server, and replace the *.dll file(s) you have recompiled
in your WebSphere "bin" directory. Restart the admin server and web server
and you are executing your modified code.
Compiling Java Code
For compiling Java code, the preferred method consists of going to the appropriate directory in the ncf, kansas,or prereqs directory tree, modifying your desired code, and invoking the compiler. For java code, you will execute the "javac" compiler which generates your "class" file (i.e. "javac UnixProcess.java" (no quotes)). For this updated class to be included for execution by WebSphere, there are two ways to do this (The author could only occasionally get the second method to work!):
The recommended method is to create a "jar" file with the associated class(es) in. Caution must be exercised in how this is done, as the packaging/class heirachy is critical. It would probably be best to find the jar your class is included in and determine any path information by either "unjarring" it or just viewing the classes within (i.e. "jar -tvf slop.jar" (no quotes)). As an example, the "UnixProcess" class is in the "admin.jar" file in the "com.ibm.ejs.sm.util.process" heirarchy, so you should go the the appropriate directory to create the jar file. In this case you must go to the "jmon" level (i.e. "c:\ws3023.sources\kansas\src\jmon" directory), and execute the jar command accordingly (i.e. "jar -cvf PQyourjarname.jar com/ibm/ejs/sm/util/process/UnixProcess.class" (no quotes)).
It is then advisable to copy your jar file to a temporary directory within WebSphere and modify your "admin.config" file's "com.ibm.ejs.sm.adminserver.classpath" parameter, (see WebSphere CLASSPATHs for details of the classpath behaviour) to include as it's first value the path to this directory. (i.e."/usr/WebSphere/AppServer/lib/temp/PQyourjarname.jar").
An alternative method is to do a normal compile, and place the compiled class file in the WebSphere "lib" directory, modify the "admin.config" file to point to your class file first in the search order (using the full path and class name, i.e. "/usr/WebSphere/AppServer/lib/yoursillyclass.class". (Note: the author only got this to work a couple of times, and quite frankly, thinks it was successful only due to magic, the correct star alignment, or his karma ran over his dogma!)
For either of the 2 above options you must restart the admin server
(and possibly the web server) with your changes and then, watch the smoke.
The following instructions, although are written with experiences from AIX and Linux, would probably work on other platforms with proper adjustments for their environments. These instructions work for the "ncf" and "kansas" source trees, and if the "prereqs" tree behaves differently, an update will be added later (as soon as I have to fix some of this code!).
(See additional information on Linux building at Configuring Linux for Compiling and Building and WebSphere Linux Build Environment )
Once you have retrieved a build and associated source files from the repository, and have installed them on your machine, you can then of course install the build on your machine using the install instructions described in other sections of this document. Within the source "ncf", "kansas", and "prereqs" directory structure resides all of the source code (java and C) for your build. To modify and test this source code, there are options on how you can proceed.
This section assumes you have done a clean and complete install of AIX or Linux and the appropriate JDK, and if necessary, the appropriate C++ environment.You should also have created a user with appropriate priviledges, are logged on as that user, and have the appropriate ncf/kansas/prereqs source files resident on your machine (preferably in the user's directory path), and you have the appropriate version of WebSphere installed. It also assumes you have your environment set up properly (for simplification, a Example Linux/AIX Build and Compile Environment Profile Script".profile" which worked for the author to setup his environments is included later, which worked for C and Java compiling). Normally all WebSphere java classes will reside in WebSphere's "lib" directory and all C/C++ code will reside in the "bin" directory.
To set up your environment, you must first modify your ".profile" file for the account you wish to do building with by adding the path to the "ncf" directory to your "PATH" environment variable, as well as assure the directories "/usr/local/bin" and the path to your Java "bin" directory (i.e. "/usr/jdk118/bin") are the first directories in your "PATH" environment variable. You must also add the following environment variables (these directories are examples, your directory paths in your environment might be different):
export JAVA_HOME=/usr/WebSphere/AppServer/java (or
wherever your JDK is)
export JANUSPATH=/home/builder/WebSphere/ncf
export LIBPATH=/usr/lib
The first variable "JAVA_HOME" must point to the directory of the java installation, and in the example above, the variable points to the java directory which ships with the build ( 3.02 only). The second variable is where you have installed your "ncf" source code directory structure.
You then have the option of doing a full build or partial build. Within the directory ncf, as well as most of the subdirectories in ncf resides some type of "makefile", which is used for making the portion of the build in that directory and it’s associated subdirectories. Regardless of where you wish to build from, you will always execute the "makeaix" or "makelinux" script (In the following instructions, the AIX platform will be assumed. If you are on Linux, substitute "makelinux" for "makeaix"). For example if you wish to rebuild the entire AIX ncf build, you would execute the "makeaix all" from the "ncf" directory. If you only wanted to rebuild the plugins for example, you would go to the appropriate directory (i.e. for the IBM HTTP server, you would go to the "/home/builder/WebSphere/ncf/src/native/plugins/apache/ibm136" directory and execute the "makeaix" with the appropriate parameter).
There are several options that can be used for executing the makefile. The typical sequence is first to clean all output files from your build directory using the "makeaix cleanall" option, then compile everything using the "makeaix all" option, and if you are rebuilding the entire build, you would typically use the "makeaix copyall" command, which will create the build in the "~/ncf/build" directory. An additional step is required for deployment of beans "makeaix deployall", and if desired, execution of the "makeaix buildImage" which will make you a "tar" file of your build, ready for installation (also there is a "makeall fromnative", don’t know what it does, and quite frankly, the author has only used the "makeaix cleanall", "makeaix all", and "makelinux all" commands).
(Note that the author did not have success using the "makelinux cleanall". He had to manually delete the *.so files and then run "makelinux all")
A typical sequence would be make the appropriate changes to the appropriate code in your directory tree (i.e. "ncf"), run the "makeaix cleanall" command from that directory, run the "makeaix all" command, and then copy your new *.class, *.so, *.a, (or whatever final output you have created) to the appropriate place in your working WebSphere installation directory, and then, dependant on what part you replaced, restart the webserver, restart WebSphere, and test your changes. If you have changed a *.so file, you should run the "/usr/sbin" "slibclean" command to clean up your libraries before starting the webserver or WebSphere.
For compiling Java code, the preferred method consists of going to the appropriate directory in the ncf, kansas,or prereqs directory tree, modifying your desired code, and invoking the compiler. For java code, you will execute the "javac" compiler which generates your "class" file (i.e. "javac UnixProcess.java" (no quotes)). For this updated class to be included for execution by WebSphere, there are two ways to do this (The author could only occasionally get the second method to work!):
The recommended method is to create a "jar" file with the associated class(es) in. Caution must be exercised in how this is done, as the packaging/class heirachy is critical. It would probably be best to find the jar your class is included in and determine any path information by either "unjarring" it or just viewing the classes within (i.e. "jar -tvf slop.jar" (no quotes)). As an example, the "UnixProcess" class is in the "admin.jar" file in the "com.ibm.ejs.sm.util.process" heirarchy, so you should go the the appropriate directory to create the jar file. In this case you must go to the "jmon" level (i.e. "/home/builder/WebSphere/kansas/src/jmon" directory), and execute the jar command accordingly (i.e. "jar -cvf PQyourjarname.jar com/ibm/ejs/sm/util/process/UnixProcess.class" (no quotes)).
It is then advisable to copy your jar file to a temporary directory within WebSphere and modify your "admin.config" file's "com.ibm.ejs.sm.adminserver.classpath" parameter to include as it's first value the path to this directory (i.e."/usr/WebSphere/AppServer/lib/temp/PQyourjarname.jar"), and possibly your "startupServer.sh" script (see WebSphere CLASSPATHs for details of the classpath behaviour) to include as it's first value the path to this directory.
An alternative method is to do a normal compile, and place the
compiled class file in the WebSphere "lib" directory, modify the "admin.config"
file to point to your class file first in the search order (using the full
path and class name, i.e. "/usr/WebSphere/AppServer/lib/yoursillyclass.class",
and possibly your "startupServer.sh" script (see WebSphere
CLASSPATHs for details of the classpath behaviour) to include
as it's first value the path to this directory.. (Note: the author
only got this to work a couple of times, and quite frankly, thinks it was
successful only due to magic, the correct star alignment, or his karma
ran over his dogma!)
Regardless of which of the above methods you use, on the Unix platforms
(Unix, AIX, Linux, etc), change the copied files (*.jar, *.so, *.a, *.class,
*.dll) file(s) owner and group to the appropriate values (i.e. "chown root
yourfilename", and "chgrp system yourfilename") and modify the file mode
to include "read" and "execute" (i.e. "chmod +xr yourfilename").On the
AIX platform, type "slibclean" at the prompt to clean up the previous library
files. For either of the 2 above methods you must restart the admin server
(and possibly the web server) with your changes, then, watch the
smoke.
Example Linux/AIX Build and Compile Environment Profile Script
# This file is the ".profile" file which is used for compilation
of
# "C" and "Java" code for the WebSphere platform on linux.
With slight
# modifications this can be used for 3.02x and 3.5x development.
# Choose one of the following two assignments to comment out
# PATH Environment assignment for Linux (Red Hat)
PATH=/usr/jdk118/bin:/usr/local/bin:/opt/IBMWebAS/lib:/usr/bin:/usr/X11R6/bin:/bin:/opt/gnome/bin:/opt/kde/bin:/usr/openwin/bin:/usr/lib/pgsql/bin:.
# PATH Environment assignment for AIX
PATH=/usr/local/bin:/usr/bin:/usr/ibmcxx/bin:/etc:/usr/sbin:/usr/ucb:/usr/bin/X11:/sbin:
ENV=.kshrc
LIBPATH=/usr/lib
# Dependant upon which version of WebSphere you are attempting
# to compile, the java environment will be different. WS302
# needs /usr/jdk_base and WS35 needs /usr/WebSphere/AppServer/java
#
# Also JANUSPATH needs to point to the correct directory wherein
# resides the associated build source tree
#JAVA_HOME=/home/aixsw/jdk/jdk1.2.2
#JAVA_HOME=/usr/WebSphere/AppServer/java
JAVA_HOME=/usr/jdk118
#JANUSPATH=/home/builder/WebSphere3.5/ncf
JANUSPATH=/home/builder/WebSphere/ncf
PATH=$PATH:$JANUSPATH
# Added the following to allow java WebSphere source code compilation
# For AIX, define the following "CLASSPATH". Substitute the appropriate
directory(s) for "yourbuilddirectory"
CLASSPATH=/home/yourbuilddirectory/WebSphere/kansas/src/jmon
# For Linux, you must have the appropriate version os WebSphere installed
and
# make the following assignments to the CLASSPATH environment variable
# Script to add all WebSphere *.jar files to CLASSPATH for Linux (Red
Hat).
JARS_AT=/opt/IBMWebAS/lib
TEMP_AT=$CLASSPATH
CLASSPATH=$CLASSPATH:$JARS_AT/BEXruntime.jar
CLASSPATH=$CLASSPATH:$JARS_AT/admin.jar
CLASSPATH=$CLASSPATH:$JARS_AT/chart.jar
CLASSPATH=$CLASSPATH:$JARS_AT/console.jar
CLASSPATH=$CLASSPATH:$JARS_AT/databeans.jar
CLASSPATH=$CLASSPATH:$JARS_AT/deployTool.jar
CLASSPATH=$CLASSPATH:$JARS_AT/dertrjrt.jar
CLASSPATH=$CLASSPATH:$JARS_AT/developBEX.jar
CLASSPATH=$CLASSPATH:$JARS_AT/developEJB.jar
CLASSPATH=$CLASSPATH:$JARS_AT/ejs.jar
CLASSPATH=$CLASSPATH:$JARS_AT/ejscp.jar
CLASSPATH=$CLASSPATH:$JARS_AT/ibmjndi.jar
CLASSPATH=$CLASSPATH:$JARS_AT/ibmwebas.jar
CLASSPATH=$CLASSPATH:$JARS_AT/iioprt.jar
CLASSPATH=$CLASSPATH:$JARS_AT/iioptools.jar
CLASSPATH=$CLASSPATH:$JARS_AT/ivbjfaceall.jar
CLASSPATH=$CLASSPATH:$JARS_AT/jacl.jar
CLASSPATH=$CLASSPATH:$JARS_AT/jetace.jar
CLASSPATH=$CLASSPATH:$JARS_AT/jskit-cfwk.jar
CLASSPATH=$CLASSPATH:$JARS_AT/jsp10.jar
CLASSPATH=$CLASSPATH:$JARS_AT/lotusxsl.jar
CLASSPATH=$CLASSPATH:$JARS_AT/ns.jar
CLASSPATH=$CLASSPATH:$JARS_AT/repository.jar
CLASSPATH=$CLASSPATH:$JARS_AT/servlet.jar
CLASSPATH=$CLASSPATH:$JARS_AT/servletconsole.jar
CLASSPATH=$CLASSPATH:$JARS_AT/sslight.jar
CLASSPATH=$CLASSPATH:$JARS_AT/swingall.jar
CLASSPATH=$CLASSPATH:$JARS_AT/tasks.jar
CLASSPATH=$CLASSPATH:$JARS_AT/tcljava.jar
CLASSPATH=$CLASSPATH:$JARS_AT/ujc.jar
CLASSPATH=$CLASSPATH:$JARS_AT/vaprt.jar
CLASSPATH=$CLASSPATH:$JARS_AT/webtlsrn.jar
CLASSPATH=$CLASSPATH:$JARS_AT/x509v1.jar
CLASSPATH=$CLASSPATH:$JARS_AT/xml4j.jar
CLASSPATH=$CLASSPATH:$JARS_AT/xml4j_1_1_16.jar
CLASSPATH=$CLASSPATH:$TEMP_AT
export PATH ENV LIBPATH JAVA_HOME JANUSPATH CLASSPATH
There are several "classpaths" within WebSphere, as well as differences in their usage.
On the Unix platforms, the admin server is started using the "startupServer.sh" script, which has within it the classpath used for the admin server. When an application server is started it uses the classpath defined in the "admin.config" configuration file.
On the NT platforms, the admin server and application servers use the classpath defined in the "admin.config" file.
Within the "~/bin/debug" directory on all platforms resides several
script/bat files which are used for specific environments, most noteably
the "adminserver.sh", which runs the admin server (without the associated
"nanny" process on the Unix platforms), and uses the classpath defined
within the script/bat file.
More details:
Here is a list and description of each and how they are used.
Administrative server classpath
This classpath is specified in the admin.config file for your WebSphere
instance. Entries added to this classpath affect the administrative
server process and all application server processes. The admin.config is
in [WAS_INSTALL_ROOT]\bin\admin.config.
Dependent Classpath
The dependent classpath is specified for the node (admin server instance,
generally the system name). It is specified via the Administrative
Console from the Topology view. The purpose of the dependent classpath
is to indicate to WebSphere where to find classes on which your enterprise
beans are dependent at deployment (Create Enterprise Bean) time.
It is never used at runtime to locate classes.
Uses:
Use the Dependent Classpath to indicate (to WebSphere's Create Enterprise
Bean function) the location of any class on which your enterprise beans
are dependent (e.g. helper classes, additional VAJ classes) which are not
in the jar file containing your beans. This node property is not
used by the application server when running your enterprise beans.
See Application Server Classpath for information on specifying the location
of helper classes during runtime of your beans.
Application Server Classpath
For enterprise beans, the application server classpath is used to specify
the location of helper classes which are not in the enterprise beans' jar
file.
For servlets, the application server classpath is used to specify the location of any helper classes which are not in the Web Application Classpath (see below), and for any servlet helper classes which use JNI (classes which invoke native methods (JNI) must be specified on the Application Server Classpath so that they are loaded by the System Class Loader and not the WebSphere classloader).
You specify the Application Server Classpath for an application server
via the Administrative Console.
1. From the Topology page, select your Application Server (e.g.
if you have the sample configuration installed, Default Server is one application
server).
2. In the right-hand pane of the console, the properties for
the application server are displayed.
3. Add -classpath paths to the Command Line Argumentsproperty.
The format of paths is the same as if you
were specify the -classpath option on the java command for your platform.
4. Click the Apply button. Wait until this completes successfully.
5. Start (restart) your application server
Note: This prepends the specified -classpath to the classpath that WebSphere sets for the application server when it start the application server process.
Note: You should NEVER use the Environment property to specify the CLASSPATH environment variable for an application server.
Uses:
To specify enterprise bean helper classes that are not in the
EJB jar file.
To specify servlet helper classes that are not specified in the
Web Application Classpath
To specify servlets or servlet helper classes which use JNI
Web Application Classpath
This is also known as the Reloadable Classpath. This is where
you would typically specify any servlets and servlet/JSP helper classes
for your web application. Classes found here will be loaded by the
WebSphere class loader. The purpose of this classpath is to allow
WebSphere to automatically reload classes when they change.
The Web Application Classpath can be found (and modified) via the Administrative Console from the Topology page.
1. Select your Web Application (you will need to expand your
application server and servlet engine tree items).
2. In the right-hand pane of the console, the properties for
the web application are displayed.
3. Select the Advanced tab of the properties
4. Enter any paths you wish to be part of the web application
classpath, one entry per line
5. Click the Apply button. Wait until this completes successfully.
6. Start (restart) the web application (right-click the web
application and select Restart Web App).
Uses:
Enables the WebSphere servlet/JSP class loaders to find servlets and
servlet helper classes and reload them when
they change (without having to end the application server). This
is the typical way to specify your servlet classes and
servlet helper classes.
Creating Model/Clones Using OSE Remote
This section describes the process of creating 2 clones on one machine which has an admin server on it, and a web server configured on a second machine. The clones will be created using the default server as the model, and the model/clones are not configured/enabled for security.
Install and start the WebSphere admin server and admin console on the machine which will contain the model and clones, and install WebSphere on the machine which will contain the web server(s), and include the appropriate web server plugin(s).
From the "Topology" tab of the model/clone machine, scope down to and
select the "default_host" (or whatever virtual host you wish to use), select
the "Advanced" tab, and in the "Host Aliases" section, add all aliases
for your web server machine (i.e. "id10t", "id10t.raleigh.ibm.com", and/or
"9.27.62.175", etc.). Scope down to the default server's servlet
engine, select it, and go to the "Advanced" tab for the engine. In the
"Queue Type" field, select "OSE". Click on the "Settings" button
and in the "Edit Servlet Engine Transport" pop-up, select "Inet Sockets"
and then click "OK". In the engine window click "Apply".
Creating the Model and Clones
From the "Topology" tab of the model/clone machine, scope down to and select the the Default Server, right click , go to "Create", and select "Model". In the "Clone Properties" pop-up window, enter a name for your model, and check the box "Recursively Model all instances under the Server", then click on "Create". Once this is done your model should appear in the topology.
Select your model, click on the "Advanced" tab, and select your "Workload Management selection policy" (the default is "Round Robin prefer local"). If you change the policy then click "Apply". Select your model in the topology frame and right click, then select "Create" "Clone". In the "Clone Parent" pop-up, scope down to your node, select it, give your clone a name and click "Create". Repeat for each additional clone you desire.
To start your clones you can start the model by selecting it and clicking "Start", or start your clones individually.
After you start the model/clones, within a few minutes a "refresh" of
the set of "*.properties" files will be generated in your WebSphere "temp"
directory (i.e. "c:\WebSphere\AppServer\temp"). You must then ftp
the "queues.properties", "rules.properties", and "vhosts.properties" to
this same directory on your webserver machine (i.e. "/usr/WebSphere/AppServer/temp")
(backing the originals up would be smart!). You must modify the "queues.properties"
files as follows:
"queues.properties" Modifications
For each clone you have created (or those you wish to use for this machine), add a line in this file "ose.srvgrp.ibmoselink.clone1.host=yourclonemachinename", where "yourclonemachinename" is the number of the model/clone machine (i.e. "ose.srvgrp.ibmoselink.clone1.host=shayden"), and repeat for each clone, incrementing the clone number. The following are the before and after version of the file.
(BEFORE)
#IBM WebSphere Plugin Communication Queues
#Tue Jan 09 18:41:52 EST 2001
ose.srvgrp.ibmoselink.clonescount=2
ose.srvgrp=ibmoselink
ose.srvgrp.ibmoselink.type=FASTLINK
ose.srvgrp.ibmoselink.clone2.port=8112
ose.srvgrp.ibmoselink.clone1.port=8111
ose.srvgrp.ibmoselink.clone2.type=remote
ose.srvgrp.ibmoselink.clone1.type=remote
(AFTER (bold lines were added))
#IBM WebSphere Plugin Communication Queues
#Tue Jan 09 18:41:52 EST 2001
ose.srvgrp.ibmoselink.clonescount=2
ose.srvgrp=ibmoselink
ose.srvgrp.ibmoselink.type=FASTLINK
ose.srvgrp.ibmoselink.clone1.host=shayden
ose.srvgrp.ibmoselink.clone2.host=shayden
ose.srvgrp.ibmoselink.clone2.port=8112
ose.srvgrp.ibmoselink.clone1.port=8111
ose.srvgrp.ibmoselink.clone2.type=remote
ose.srvgrp.ibmoselink.clone1.type=remote
"bootstrap.properties" Modifications
You must also make changes to your "bootstrap.properties" file on the web server machine to reflect the clones. For each clone you have created (or those you wish to use for this machine), add 3 lines in this file for each clone, "ose.srvgrp.ibmoselink.clone1.port=yourcloneportnumber", where "yourcloneportnumber" is the number of your clone (i.e. "ose.srvgrp.ibmoselink.clone1.port=8111"), "ose.srvgrp.ibmoselink.clone1.type=yourclonetype", where "yourclonetype" is the type of your clone (i.e. "ose.srvgrp.ibmoselink.clone1.type=remote"), "ose.srvgrp.ibmoselink.clone1.host=yourclonehostname", where "yourclonehostname" is the number of your clone (i.e. "ose.srvgrp.ibmoselink.clone1.host=shayden"), and repeat for each clone. Adjust the "clonescount" parameter accordingly.
The following are the before and after version of the file, but only
the last few lines of the file needs modying and are therefore the only
part of this file shown in these examples.
(BEFORE)
ose.adminqueue=ibmappserve
ose.max.conncurrency=1
ose.srvgrp.ibmappserve.type=FASTLINK
ose.srvgrp.ibmappserve.clonescount=1
ose.srvgrp.ibmappserve.clone1.port=8081
ose.srvgrp.ibmappserve.clone1.type=local
ose.srvgrp.ibmappserve.clone1.host=localhost
ose.mode=out
(AFTER (bold lines added/changed))
ose.adminqueue=ibmappserve
ose.max.conncurrency=1
ose.srvgrp.ibmappserve.type=FASTLINK
ose.srvgrp.ibmappserve.clonescount=2
# The following three lines are comments now
#ose.srvgrp.ibmappserve.clone1.port=8081
#ose.srvgrp.ibmappserve.clone1.type=remote
#ose.srvgrp.ibmappserve.clone1.host=shayden
ose.srvgrp.ibmappserve.clone1.port=8111
ose.srvgrp.ibmappserve.clone1.type=remote
ose.srvgrp.ibmappserve.clone1.host=shayden
ose.srvgrp.ibmappserve.clone2.port=8112
ose.srvgrp.ibmappserve.clone2.type=remote
ose.srvgrp.ibmappserve.clone2.host=shayden
ose.mode=out
Startup and Use of Web Server(s)
At this point your admin server machine which contains the model and clones has been started, your web server machine has been propertly configured, you can then start the web server(s) on your web server machine. You should then be able to execute the "showCfg" servlet from any machine's browser (i.e. from the machine "lucy4" browser using "http://id10t/webapp/examples/showCfg") see that the "hostname" reflects the host of the machine running the clone (i.e. "hostname: shayden.raleigh.ibm.com"), that the type is "remote", the port reflects the correct port for the clone (i.e. "port: 8111), and the "cloneIndex" reflects the clone number (i.e. "cloneIndex: 1"). When you reload the page you should see the behaviour of the round robin "Workload management selection policy", in this case alternating between clones 1 and 2.
Note that if you make any changes to your model/clone (or practically
ANYTHING) configuration on your admin server machine, you will refresh
your properties files on that machine. You will then need to copy the updated
files to your web server machine and make the appropriate changes described
above.
Debugging
Utility (DBX) on Solaris and AIX
DBX is a AIX/Solaris debugger and is used for troubleshooting certain types of problems on the AIX platform. There is some likelihood the debugger is already installed on your system, but if not, the following instructions describe how to install it.
Mount the AIX cd # 1 (see instructions above), and start "smitty", select the following:
Software Installation and Maintentance
Install and Update Software
Install Software Bundle (Easy Install)
(press PF4 for a list of options, one should be the cd, select
it and press "Enter")
(in the "Select a Fileset Bundle" window, select "App_Dev")
Running DBX on AIX
Running DBX on Solaris
Example DBX
Trace Sessions
You must invoke dbx from an AIX window using the command "dbx –a (process id) (executable name)", for (i.e, "dbx –a 6969 httpd"). The process id and executable name are the AIX process id for the executable, and the actual executable itself. Once you type in this command, dbx will be invoked and after it’s initialization and reading in the appropriate information, will respond with a "dbx" prompt. At this point type in "cont" to continue dbx execution. When the process "httpd" in this example does a segmentation fault, control will be returned to dbx. At this point you can display the last set of functions/methods called before the segmentation fault by issuing the command "where" at the dbx prompt.
Or if debugging a core file, execute dbx in the directory the core file
is in, and give dbx as a parameter the java path that is executing (i.e.
"dbx /usr/java1.1/bin/sparc/native_threads/java"). DBX might display
new features and commands and read some libraries. Type "where" and you
(might) get a list of functions/methods executed while your application
was dying and generating the core file.
On Solaris the DBX utility comes with the Workshop Pro utility package (This section assumes you have installed this package and associated license). To invoke dbx on Solaris you need only type the process id (i.e., determine the process id of the "http" process (e.g. 16566) and use it in the command line, "/opt/SUNWspro/bin/dbx - 16566"). Once you type in this command, dbx will be invoked and after it’s initialization and reading in the appropriate information, will respond with a "dbx" prompt. At this point type in "cont" to continue dbx execution. When the process "httpd" in this example does a segmentation fault, control will be returned to dbx. At this point you can display the last set of functions/methods called before the segmentation fault by issuing the command "where" at the dbx prompt.
Or if debugging a core file, execute dbx in the directory the core file
is in, and give dbx as a parameter the java path that is executing (i.e.
"/opt/SUNWspro/bin/dbx /usr/java1.1/bin/sparc/native_threads/java").
DBX might display new features and commands and read some libraries. Type
"where" and you (might) get a list of functions/methods executed while
your application was dying and generating the core file.
Begin by starting up dbx monitoring a process (this is an AIX session).
AIX/DBX prompts are in Blue, andcommands typed
in are in Green:
Example 1
[root@id10t:/usr/HTTPServer/bin]--dbx
-a 26288 httpd
Waiting to attach to process 26288 ...
Successfully attached to httpd.
Type 'help' for help.
reading symbolic information ...warning: no source compiled with -g
stopped in read at 0xd016fdf0 ($t1)
0xd016fdf0 (read+0x114) 80410014
lwz r2,0x14(r1)
(dbx) cont
Segmentation fault in _doprnt at 0xd017a250 ($t1)
0xd017a250 (_doprnt+0x18) 88030000
lbz r0,0x0(r3)
(dbx) where
_doprnt(0x5a5a5ca6, 0x2ff21324, 0x2ff20280) at 0xd017a250
vsprintf(??, ??, ??) at 0xd0179ac8
logger.ose_log(0x200806c0, 0x1, 0x5a5a5ca6, 0x2005fbc0, 0x50, 0x20060010,
0x9, 0x5a5a5a5a) at 0xd170c3bc
sec_is_request_protected(0x20050978, 0x2ff21848, 0x2ff21830, 0x20050988,
0x200806c0) at 0xd1715e60
ws_is_request_protected(0x20050988, 0x2ff21848, 0x2ff21830) at 0xd170ab90
as_check_user_id(0x2005e390) at 0xd16dc1e4
run_method(??, ??, ??) at 0x100223d4
process_request_internal(??) at 0x10048178
ap_process_request(??) at 0x100496d0
child_main(??) at 0x10002fc0
make_child(??, ??, ??) at 0x100029b4
perform_idle_server_maintenance() at 0x10002684
standalone_main(??, ??) at 0x10002024
main(??, ??) at 0x100018cc
(dbx)
Example 2
[root@id10t:/usr/HTTPServer/bin]--dbx
-a 26288 httpd
Waiting to attach to process 26288 ...
Successfully attached to httpd.
Type 'help' for help.
reading symbolic information ...warning: no source compiled with -g
stopped in child_main at 0x10002d00 ($t1)
0x10002d00 (child_main+0x2f0) 80410014
lwz r2,0x14(r1)
(dbx) cont
Segmentation fault in ws_vhost_find at 0xd4149284 ($t1)
0xd4149284 (ws_vhost_find+0x88) 80840004
lwz r4,0x4(r4)
(dbx) where
ws_vhost_find(0x7c03232e, 0x2ff213b4, 0x5a5a5a5a, 0x2ff213b0, 0x200806c0)
at 0xd4149284
sysmgmt_vhost_for_aliasport(0x854a0, 0x2ff213b4, 0x5a5a5a5a, 0x2ff213b0,
0x200806c0) at 0xd4145ef8
sysmgmt_queue_for_uri(0x854a0, 0x2005f8c0, 0x50, 0x2005f2a8, 0x2ff21820,
0x200806c0) at 0xd41457ec
(dbx) quit
There are several types of logs that can be captured in running WebSphere, and although they can be switched on relatively easy, they do require resources and can affect performance of your applications. A typical procedure when debugging classses, methods, and/or functions would be to add trace statements to the code that perhaps displays some variables to be output for subsequent analysis of the associated log files.
WebSphere
Plugin Trace
WebSphere
Tracing
There are two types of logging that can be enabled for tracing web server plugins in the system. On the "client side", each webserver plugin has two points at which logging can be done. In the WebSphere "properties" directory there is a file "bootstrap.properties", which has two variables "ose.native.log.level" and "ose.plugin.log.level" which by default are only set to log warnings and errors. If you wish to enable trace and information logs, you must include these parameters in the values for these variables (i.e. "ose.native.log.level=TRACE|INFORM|ERROR|WARNING"). Trace logs from this configuration will appear in your WebSphere log directory preceded by a "trace.log.(webserver type)" prefix followed by a date/time stamp of when the web server was started (i.e "trace.log.ibmhttp.Wed=Jun-28-09.29.11-2000").
On the WebSphere "server side" of the plugins you can turn on logging
from the Administrators Console by selecting your servlet engine, and in
the "Advanced" tab, select "Settings", and in the pop-up, select the level
of logging you wish to capture. These logs will also reside in the WebSphere
logs directory with the filename "native.log.was-oop(with a date/timestamp").
In the topology, under the appropriate node, select the appropriate
Application Server, select the "Advanced" tab, and in the "Trace Specification",
enter the value for the level of tracing you wish to record (i.e."com.ibm.*=all=enabled"
(with no quotes)), and all trace code in this level on down will be recorded
to the console, or you can specify a "Trace output file" you wish your
tracing to be written to.
There are occasions when WebSphere needs it's databases cleaned (i.e., lockup, a barage of exceptions, etc.). These instructions describe the process of clearing out the database and setting the heapsize accordingly. If you do this process it will be necessary to modify your "admin.config" file to set the "install.initial.config" to "true" if you wish to have the sample "default" configuration installed (i.e. default host, default application, etc.) before you start your admin server. In the following examples, text in parenthesis are comments.
DB2 Database Cleaning
Oracle Database
Cleaning
The DB2 database is first deleted ("dropped"), then recreated, and the
heapsize is changed. This can be done manually or (if in AIX/Linux) an
alias or script can be created to execute the following 3 db2 commands.
Choose the platform/method you have from the following options.
If using Windows NT or 2000, from the "Start", "Programs", "IBM DB2", select the "Command Line Processor". At the command prompt issue the following commands:
drop database was
create database was
update db config for was using applheapsz 256
(OR)
In AIX/Linux, you can create and execute a shell script with the following contents:
db2 drop database was
db2 create database was
db2 update db config for was using applheapsz 256
(OR)
In AIX/Linux, you can create an alias, which can be assigned at the command line or in your ".profile" is as follows:
alias dbw="db2 drop database was;db2 create database was;db2 update
db config for was using applheapsz 256"
The following are two different methods which clean up the ejsadmin database/schema used by WebSphere:
While logged on as "root" id, perform the following actions:
(METHOD 1)
su – oracle
sqlplus system/manager
drop user ejsadmin cascade;
create user ejsadmin identified by yourpassword;
("yourpassword" = your ejsadmin password)
grant connect,resource,dba to ejsadmin;
(OR)
(METHOD 2)
su – oracle
sqlplus
(Login as "ejsadmin" using password
"ejsadmin")
(note, the following path is an example, you can store the script anywhere)
SQL start /export/home/yourdirectory/scripts/dropTbl.sql
quit
Note that in the following script, the last line lists any tables that still exist for this user. If the execution of this script displays that there are additional tables that exist, they must be added to this script and/or subsequently manually dropped for a complete cleaning to occur..
The above sql script has in it the following commands:
drop table APPL_TABLE;
drop table APPSECURITY_TABLE;
drop table BEAN_TABLE;
drop table BINDINGBEANTBL;
drop table CONTAINER_TABLE;
drop table CONTEXTBEANTBL;
drop table DATASOURCE_TABLE;
drop table EBSECURITY_TABLE;
drop table EJB_SERVER_TABLE;
drop table LIVE_OBJ_GRP_TBL;
drop table LIVE_OBJ_TABLE;
drop table LTPACONFIG_TABLE;
drop table MG_TABLE;
drop table MODEL_TABLE;
drop table NODE_TABLE;
drop table OBJECT_TABLE;
drop table OLT_TBL;
drop table PROPERTYBEANTBL;
drop table REL_DEFN_TABLE;
drop table REL_INSTANCE_TABLE;
drop table SECURITYCFG_TABLE;
drop table SERVER_TABLE;
drop table SESSION_MGR_TABLE;
drop table LOCK_TABLE;
drop table SE_TABLE;
drop table SRVLT_ENG_TBL;
drop table SRVLT_GRP_TBL;
drop table SRVLT_REDIRECT_TBL;
drop table SRVLT_TBL;
drop table TYPE_TABLE;
drop table URI_TBL;
drop table USER_PROFILE_TABLE;
drop table VIRTUAL_HOST_TBL;
drop table JDBCDRIV_TABLE;
drop table REPOSITORY_LOCK_TABLE;
select table_name from all_tables where owner = 'EJSADMIN';
The tool can be obtained from the machine "id10t" using the "guest" account (password also "guest"), in the directory "/home/aixsw/ws.tools/" (alternatively it can be retrieved from http://water.raleigh.ibm.com/aktools/download/index.html , but you must uncompress the file first). FTP and untar the file "AKtools_AIX_tar.tar" on your machine.
The web site http://water.raleigh.ibm.com/aktools/docs/index.html has instructions/documentation on the running of this utility.
A simple configuration will be to copy the existing configuration file "akstress.acf" which is in the "AKtools/config" directory to the "AKtools/bin" directory, and then modify the copy to point to your server (modify the "Host" field, i.e. "Host: id10t", and to serve your page modify the "URI", i.e. "URI /servlet/hello"), modify the "ResultCode" (which should be "200") for a normal retrieved page (the "akstress.acf" file contains a definitions of the various settings which affect the stress run). You can also modify the location/name of the "ErrorLog" and "AccessLog" variables. Other values you might change include "Threads" and "TotalPageRequests".
After making the above modifications to the copy of the "akstress.acf" file in the "AKtools/bin" directory, from that directory execute the akstress command to invoke the utility:
(i.e. "./akstress -host id10t.raleigh.ibm.com:80
-config akstress.acf" (no quotes)).
Adobe Forms Data Format Installation
Get the files from http://partners.adobe.com/asn/developer/acrosdk/forms.html
and uncompress and untar them. In the directory created ("SSOLRS.install"),
execute the install script (i.e. "./INSTALL"). Foller the instructions,
and use the defaults.
Unix/Linux/AIX Operating System Commands/Utilities
The commands/utilities listed here are common for most Unix environments. However, some required the installation of various utilities and subsystems. All commands have worked before, so if they dont work on your system, you are on your own.
Common
Unix/Linux/AIX Commands
Common
File Transfer Protocol (FTP) Commands
System
Management/Info/Miscellaneous
Linux
Specific Operating System Commands
License
User Manager
MQSERIES
Commands
MISC
vi Commands
Using
the Systems Management Tool for Uninstalling Software
Mounting
a CD on AIX
Using
The Systems Management Tool To Increase File Sizes
Using
The Systems Management Tool to Add Users and Groups
Creating
a new user using "smit" or "smitty"
Creating a new group using "smit" or "smitty"
Miscellaneous
(Perhaps Useful) Web Sites
Common Unix/Linux/AIX Commands
Command Action
nslookup (ip or nodename)
Looks up node name or ip address on net
su
Log in as root (will be queried for password)
find / -name (filename)
Finds (filename) looking from root (/)
find . -name (filename)
Finds (filename) looking from current directory
file (filename)
Displays what type of file (filename) is
find (path/subs) -name (filename) Finds (filename)
starting at (path/subs)
* domainname
Returns domain name
* ypmatch (nodename) hosts
Returns IP address and alias'
uname -a
Returns node, uxversion, machine type, etc.
hostname
Returns the domain name of this server
whoami
Returns your user id
shutdown -r now
Shut the box down, reboot!
chmod -644 (filename)
Changes file to read by everyone
chown (owner) (directory/filename) Change owner of filename/directoryname
chgrp (group)(directory/filename) Change group
of filename/directoryname
* bdf
Displays disk space
rm -r (directory)
Removes directory and all subdirectories
rm -rf (directory)
Removes directory and all subdirectories (no prompts!)
tar -cvf (file.tar) ./(directoryname) Tar's a directory
and all subs
usr/openwin/bin/xhost +
Tells xserver to accept x requests from any host
ls -aCF
Lists files (inc. "."'s)
unset (environment variable)
Unsets an environment variable
du -s (directory name)
Lists space used in directory and subs
du –k .
Lists space used in directory and subs (in kilobytes)
df -k
Lists space on disks in kbytes
* df -b
Lists available space on current disk in blocks
df .
Do df on current directory/filesystem
finger
Queries who is on and is doing what
tar -cvf newtarfile directory
Directory/file to tar
tar -xvf (filename.tar) &
Untar a tar file
tar cvf /dev/rmt/0mn (filename)
Tar my file to mag tape with rewind
tar cvf /dev/rmt/0m (filename)
Tar with no rewind
tar tf
Look at next tar file on tar tape
tar tvf (filename) | grep (string)
Search tar file (filename) for file (string)
tar -cf - (directory/file) | rsh (nodename) 'cd vob1; cat (filename)
Tars from another machine
. (filename)
To execute script file
jobs
Displays jobs running invoked from current shell
mount -v 'cdrfx' -r /dev/cd0 /usr/cdrom/ics
Mount cdrom on AIX
ls | wc –l
Lists number of files in a directory
? mount -F nfs (nodename):/cdrom/(directoryname) /cdrom
Mount a cdrom on diff. machine
? umount /cdrom
Unmount cdrom
sty
Sets terminal characteristics (ie. stty erase (backspace character)
nm libcrap.so
Displays symbol map and other info for a *.so or *.a file
Piping
Command/script 2 afilename Pipes output of the execution of a command to afilename
$$ Korn shell
ksh
Go to Korn shell
* whence (program name)
Tells you directory the program exe. is
export (env var)=(value)
Sets environment variable in Korn/Borne Shell
$$ Misc NSF stuff:
* /etc/services
Has list of ports on workstation
/etc/hosts
Has IP addr. (resident machine & ones you can acc.)
* /etc/nodename
Has hostname in it (i.e. greenk.charlotte.ibm.com)
$$ Netscape stuff
* /usr/local/lib/netscape
(expects to go)
* /usr/contrib/bin
(put executables in for Navigator)
* Dependant upon your particular installation, this does
not work on IBM AIX
Common File Transfer Protocol (FTP) Commands
cd directorypath
Change to the associated directory on the destination machine
del filename
Delete the associated filename
lcd directorypath
Change to the associated local directory
ascii
Configure ftp to copy an ascii/text file
bin
Configure ftp to copy a binary file
pwd
Print the name of the working directory
prompt
Turn off prompting for multiple files to be transferred
put
Put a file on the destination machine
get
Get a file from the destination machine
(the following commands can use wild cards)
mget *.*
Get all files in the destination machine's current directory
mput *.*
Put all files in the current local machine to the destination machine
Linux Specific Operating System Commands
rpm -q softwarepackagename
Is the software package "softwarepackagename" installed on system?
rpm -qa
Lists all installed software installed
rpm -e erasethissoftware
Uninstall the software package "erasethissoftware"
rpm -ivh installthissoftware
Install the software package "installthissoftware"
rpm -q IBMWebAS
(shows the rpm package name)
rpm -qi IBMWebAS
(shows lotsa details about the package, location, description, etc)
rpm --uninstall IBMWebAS
(uninstalls the package)
man rpm
(bores you to death with details and "the philosophy" of rpm)
smbclient //wasdoc0/APARS -U guest
Creates a more robust ftp connection (allows you to read a file)
System
Management/Info/Miscellaneous
smit
Systems management utility (GUI)
smitty
Text version of Systems Mgt utility
info
Documentation utility
xcalc
Calculator
lsvg rootvg
Lists total free disk space
lslpp –l
Lists software component versions
lslpp –l db2*
Lists S/W component requested (i.e. DB2)
lsvg
List volume groups
lspv
List physical groups
lsfs
Lists file system
startsrc –g nfs
Starts NFS process
shutdown –Fr
Shuts down/reboots system (DANGEROUS!!!!)
shutdown -r now
ulimit –a
Shows what you userid is allowed to use
vmstat
Lists virtual memory
vmstat –s
Lists swap memory
vmstat –i
Lists interrupts
vmstat 10
Runs vmstat every 10 seconds
ps aux
Lists everything about processes
uptime
Displays how long since last reboot
lsps –s
Lists paging space
svmon –G
Lists memory usage
oslevel
Lists OS level
chmon -p20
Invokes real time cpu monitor/display (top 20 processes)
tprof
AIX system monitoring tool
nslookup (machinename)
Lookup IP address value/existence for machine/domain name
jar –tvf (jar name)
List the classes within a Java jar file (i.e. "jar –tvf servlet.jar")
fsck
File System Check, do on reboot when machine has been shut down dis-gracefully
monitor -a
Systems monitoring tool for showing CPU usage, etc.
xload
Systems monitoring tool for graphically showing aggregate machine usage
netstat -a
Provides detail of networking configuration
$$ * If you try to log into a machine as root
from another console, and it will not let you, you can
$$ * go to the /etc/default/login
and in login command, comment out console command.
i4blt –a –n (server name)
i4cfg –list (lists subsystems)
i4blt –lp (lists licenses)
runmqsc WBANK0
Invoke MQSeries Admin tool and look at qmanager WBANK0
/usr/lpp/mqm/bin/runmqsc WBANK0 Same as above with command not
in your PATH
dis q (‘*’)
Shows all queues
runmqsc WBANK0 silly
Run runmqsc on WBANK0 and pipe all subsequent output to "silly"
(Note that output will not appear on
your display when invoking subsequent commands. When you
terminate the running of "runmqsc"
you can edit this file)
<ctrl/C
Exit runmqsc
dspmqfls -m WBANK0 -t qr ‘*’
Display all remote queues on WBANK0 queue manager
dspmqcsv WBANK0
Display status of WBANK0 command server
strmqm WBANK0
Start queue manager WBANK0
endmqm WBANK0
Stop queue manager WBANK0 (there are 2 other ways to stop a queue manager)
:1,$s/oldstring/newstring/g
Substitute oldstring with newstring from first to last lines
:%s/oldstring/newstring/g
Same as above
:1,$?/oldstring?/newstring?g
Same as above but will look for or substitute slashes "/"
:30,50s/oldstring/newstring/g
Substitute oldstring with newstring on lines 30 through 50
:set ic
Disable case sensitivity while searching for strings
:1,$s/^.........//
On every line, delete the first X characters where X = the number of "."
characters
First create a directory in your file system to refer to your CD (i.e." mkdir CD" from the root directory), and then issue the following command to mount the CD drive: "mount -v'cdrfs' -f'' -r'' /dev/cd0 /CD" (the double quotes around this command are for clarity and are not part of the command. All of the quotes IN the command are single quotes). At this point it can be accessed like any other directory.
(If the above command line type of mount does not work:)
Using "smit" or "smitty": (selections are):
System Storage Management
File System
Mount a file system
(in the pop-up, enter:)
File System Name (use list and select the highest digit drive,
i.e. "/dev/hd4"
DIRECTORY over which to mount (must have been already created,
ie. "/CD"
Type of file system "cdrfs"
Force the mount? "yes"
Remote node..... (ignore this one, leave blank)
Mount as a REMOVABLE file system? "no"
Mount as READ-ONLY system "yes"
Disallow DEVICE access via this mount? "no"
Disallow execution of SUID and sgid programs....
"no"
Then click "OK"
Using
The Systems Management Tool To Increase File Sizes
To use the "smitty" systems management tool to increase the sizes, from
the command line type "smitty". Select ""System Storage Management", the
"File Systems", then "Add/Change/Show/Delete File Systems", "Journaled
File Systems", then "Change/Show Characteristics of a Journaled File System".
Select "/home" in the pop-up and change "Size of file system" to "81920".
Use this same procedure to change "/usr" to "3809280", and "/" to 65536
(if any of these values are less than what you system is already configured
at, use the largest value for that file system).
Using The Systems Management Tool to Add Users and Groups
Creating
a new user using "smit" or "smitty": (selections are):
Security and Users
Users
Add a User
User Name es363
Home Directory /home/es363
(remaining values can be ignored)
Creating
a new group using "smit" or "smitty": (selections are):
Groups
Add a Group
Group Name es363
Administrative Group false
Group ID
User List es363
Using the Systems Management Tool for Uninstalling Software (selections are):
Software Installation and Management
Software Maintenance and Utilites
Remove Installed Software
(From this window, pressing key "F4" provides
a list of all software. Go
to the desired item(s) you wish to remove
and press "F7" to select it/them. Press
"Enter". Your items will be included
in the "Entry Fields". Tab to the next
entry "Preview only?", and press the "Tab"
key to indicate "No". Press "Enter"
and the software will be uninstalled)
(notes: The WebSphere product will normally be "IBMWebAS", and may contain several modules, dependant upon your installation.
There is some likelihood you will need to delete associated directories
dependant upon what software you are uninstalling).
Unix Stuff
WebSphere
DB2
Solaris
AIX
HTTP Server
Apache Server
NT Stuff
To determine network information from DOS command prompt (IP addresses, etc). "ipconfig /all" (no quotes).
To paste text in a dos window from a windows environment (Notepad, Wordpad, Word, etc. "ctrl/c"), go to the DOS window and do a right mouse button to paste.
To map a network drive (i.e. "scarecrow"), from Windows Explorer, select "Tools", and "Map Network Drive". In the Map Network Drive window, select "S" as drive name, and "\\scarecrow\share$", and in the "Enter Network Password", enter your id as "sapiensamericas\yourname.yourfirstletter", and your password, and click "OK" (or maybe it's "Finish").
DWB patch files from Gil will be in the "Temp File Share\GilS\Patch"
Most forms of Unix (AIX, Solaris, Linux, etc) normally employ a similar mechanism for startup. Typically, the script file /etc/inittab contains the set of operating (sub)systems which are to be started on reboot, dependant upon the level of severity. The directory "init.d" typically contains scripts which perform specific operations during startup of these subsystems. Parallel to init.d is a set of directorys which contain links to the scripts in the init.d directory. These directories each contain a set of symbolic links which point to the associated script file in "init.d". Dependant upon the first letter of the link, "K" or "S", the associated link's script will be run. Those beginning with "K" are run when shutting down the machine, and those beginning with "S" are run on (re)starting the machine. There are normally 7 different levels available for startup, depending on the severity of startup required, each level with a corresponding directory (i.e. "rc0.d", "rc1.d", through "rc6.d").
Unzipping A Single File (text files only)
From the directory your *.zip file is in, issue the command "unzip -c
yourzipfilename.zip thefullpath/yourtextfile newfilename", where "yourzipfilename.zip"
is the zip file which contains your desired file, "thefullpath" is the
full path name of your file, and "yourtextfile" is the name of the file
you wish to extract. The file will be "piped" into the file "newfilename"
(ie. "unzip -c ncf.zip ncf/src/native/ose/asout_lib/marshal_util.c marshal_util.c",
with no quotes). The resulting file will have 2 lines at the beginning
and one blank line at the end which you must delete.
To determine the version of WebSphere, Java, etc., on a installation of WebSphere with the admin and web server running as well as the default application running, use a browser to execute the "showCfg" servlet (i.e. "/http://id10t/webapp/examples/showCfg" (no quotes)).
To turn off the Java Just In Time Compiler (JIT) for a particular WebSphere Application Server, from the WebSphere Admin Console, select your application server (in WS 302x, select "Topology" and scope down to this), and in the "General" tab, in the "Command line arguments:" field, enter the value "-Djava.compiler=none" (no quotes). "Apply" this, and restart the application server.
To log verbose tracing for a particular WebSphere Application Server, from the WebSphere Admin Console, select your application server (in WS 302x, select "Topology" and scope down to this), and in the "Advanced" tab, in the "Trace specification:" field, enter the value "com.ibm.*=all=enabled" (no quotes). "Apply" this, and restart the application server.
To log only garbage collection tracing for a particular WebSphere Application Server, from the WebSphere Admin Console, select your admin server, and in the "General" tab, in the "Command line arguments:" parameter, apply the statement "-verbosegc" (no quotes). "Apply" this, and restart the application server.
To find out what version of the build you have on your machine, look in your WebSphere root directory in the "properties/com/ibm/WebSphere" at a file "product.properties".
Alternative source for builds is "wasbuild", id/pw = asteam/asteam
DB2 has absolutely no problems and your desperate attempt to look here
rather than the appropriate DB2 documentation can only be rewarded by your
frustrations at reading this, the only useful sentence in this section.
HOWEVER, some common DB2/SQL commands:
Common DB2 commands:
This section describes common DB2 commands which might be useful. This
can be done by typing
db2 at the command line prompt and entering the following commands
(minus the "db2" argument and double quotes), or the recommended method
is to issue the commands as defined below from the promt. The latter
method allows you to use the command stack. Using this method also
allows you to use the pipe to "more" or "pg". DB2 is not case sensitive,
but it is value type sensitive, i.e. if you try to change a value surrounded
by single quotes (‘), it assumes a character value, not a numeric value.
In the following commands, DB2 is very picky about the type of quotes you
use.
To use the following commands you should be logged in with the userid of the owner of the associated database (i.e. the "root" userid is the owner of the "was" DB2 database, and issue a "connect to <yourdatabase" (i.e. db2 connect to was). The commands themselves are usually case insensitive, but the data is not. All alphanumeric character data must be in single quotes, and numerical data needs no quotes.
NOTE: THE FOLLOWING COMMANDS (AND DB2 COMMANDS IN GENERAL) CAN HAVE DRAMATIC AFFECTS ON YOUR DATABASE. PROCEED WITH CAUTION WHEN ALTERING/UPDATING DATA IN YOUR DATABASE. ESPECIALLY IF YOU SHARE A DATABASE!!!!!!!!, THESE COMMANDS CAN GET YOU OSTRACIZED FROM SOCIETY.
The following commands are examples, AND ONLY EXAMPLES!
db2 ? Lists DB2 commands
db2 connect to was Required initial connection
to database to do SQL type commands
db2 disconnect was Disconnects from this
database
db2 "list database directory" Lists all databases
on this machine
db2 "list tables" Lists all tables
in the connected database
db2 "select * from ifs_fi" Display all
info for FI in ifs_fi table
db2 "select user_id from ifs_user" Displays
all users
db2 "select user_id, password, pin, logged_on from ifs_user"
To display all users/passwords/pins
db2 "update ifs_user set logged_on=’N’" To reset
all logons when a user was logged
on to an account when server
was restarted
db2 "update ifs_transfer set transfer_status=’2’" Updates
ALL transfers to "history"
db2 "update ifs_transfer set transfer_status=’2’ where user_id=’ARF1001’"
Updates for ARF1001
db2 "select logged_on from ifs_user where user_id=’ARF1001’"
Is userid ARF1001 logged on?
db2 "update ifs_user set logged_on=’N’ where user_id=’ARF1001’"
To reset an individual userid
db2 "drop database YOURDATABASENAME" Kill the associated
database
db2 "create database YOURDATABASENAME" Create the database
db2 "force applications all" Release all applications
connecting to the database you are connected to.
(The following commands describe how to export data from one database
and put it in another
database. Obviously you must be logged in as the appropriate
database owner and connected to the
appropriate database for each action, as well as the file (ARF1001.ixf)
must be readable by the
second database id/owner. You can export in other formats which
would allow manually editing of the
data, this is not described here)
db2 "connect to was"
db2 "export to ARF1001.ixf of ixf messages msgs.txt select *
from ifs_transfer where user_id=’ARF1001’"
(in the above statement, we are exporting all transfers in the
"ifs_transfer" table which have a user_id of
"ARF1001" to a file "ARF1001.ixf"" in the "ixf" format, and logging
any status/error information in
"msgs.txt". The following commands import this data into another
database)
db2 "connect to was"
db2 "import from ARF1001.ixf of ixf insert into ifs_transfer"
Import data for table "ifs_transfer" from file
"ARF1001.ixf" using "ixf"
format.
On Solaris, (maybe even AIX) you need to run ksh to be able to export display. Likely you will need to "xhost +" also.
On Solaris or AIX, to run db2 from the command line, you must be logged on as a user that is a member of all 3 db2 groups (Solaris = "db2asgrp", "db2fadm1", and "db2iadm1".
To add user on Solaris, use the Administration Tool, which is invoked
from the command line with the command "admintool". And as per previous
"tidbit", go ahead and add the three db2 groups.
When running "configassist", you cannot execute your .profile beforehand, for if you do you will get the error "Cannot find class "java/lang/Thread".
On Solaris, (maybe even AIX) you need to run ksh to be able to export display. Likely you will need to "xhost +" also.
On Solaris or AIX, to run db2 from the command line, you must be logged
on as a user that is a member of all 3 db2 groups (Solaris = "db2asgrp",
"db2fadm1", and "db2iadm1").
The Apache server comes shipped with some Linux systems (Red Hat for sure). To
Starting Apache HTTP Server:
Change directories to "/etc/rc.d/init.d" and issue the command "./httpd
start" (no quotes)
Stoppng Apache HTTP Server:
Change directories to "/etc/rc.d/init.d" and issue the command "./httpd
stop" (no quotes)
To ftp to a web server and specify the id/password in the uri, ftp://shayden@tikki69r@zircon.austin.ibm.com/
However, this still requires you to type in the password on zircon.
To change httpd admin password (from your servers directory, i.e. C:\Program
files\IBM Http Server)
type "htpasswd –c conf\admin.passwd db2admin" and you will be prompted
for new password.
Starting IBM HTTP server:
/usr/HTTPServer/bin/apachectl start
Stopping IBM HTTP server:
/usr/HTTPServer/bin/apachectl stop
Determining the Version of IBM HTTP server:
/usr/HTTPServer/bin/httpd –V (displays version of http server)
Miscellaneous
(Perhaps Useful) Web Sites
http://wsdcds01.watson.ibm.com/wsdc.nsf/guides/download.htm
(WSDC Download Guide)
http://w3-1.ibm.com/download/standardsoftware/index.html
(IBM Standard Software
http://java.sun.com/products/jdk/1.2/docs/api/overview-summary.html
(Java 2 Platform API Spec
Author’s
Notes (Read only if incredibly bored and/or you did not understand something
from previous sections!):
Steve, you had a "ioser library" problem, Expected v1.10 but loaded v1.9. Talked to Mike Morton who said to talk to Steve Krol about what version of JDK 1.1.8 patches to install. Left message, has not returned call. Tried several times to run "./startupServer.sh" but got the same error. Eventually it did start and got "Open for e-bitness in the "tracefile". However, upon starting "./adminclient.sh &" and the GUI coming up, it wasn’t doing anything to the database and the default server was not in the topology view. Also, nanny wasn’t running.
Tried to uninstall java 1.1.8 but there were prerequisites (web based systems management), which I uninstalled with smitty, then uninstalled JDK with smitty. I then went to /CD/aix in the WebSphere CD and executed "./install.sh", which (text based) discovered no jdk was installed and asked if I wanted to install from the distribution disk (duh "yes"). I then tried "./startupServer.sh &" again. Got "Open for e-bitness" in tracefile again. This time "ps –ef |grep java" showed the AdminServer and nanny processes running. Waited a month. Client came up, default_host was there, but no jvm/etc.
Uninstalled websphere and deleted directory. Then reinstalled websphere with different db2 acct/password, and everything started ok.
NOTES on PMR87850
Installed adobe API, added FD*.jar directory to NT Service classpath,
servlet was in servlets directory. Had to compile test file in servlets
directory and copy to servlets/com/wiredscholoar/bizlogic/loans.
Had to add C:/Program Files/Adobe/AcrobatSDK/Acro4SDK-R3/Fdftk/Lib/Java/FdfTk.jar
to Websphere defaultApp classpath. Created /Program Files/Adobe/AdobeDLLs
and copied dll files from adobe install to this directory and added this
to
Booted up new x86 machine with Winders 98, got:
"CMOS Settings:"
"What do you want to do with the CMOS settings"
(A) Accept
(I) Ignore
(Q) Quit
I chose "A" to accept.
"Master Boot Record"
ditto, chose "A" accept
Boot Sector OK
Bios System File OK
Dos System File OK
Dos Shell File OK
Would you like to save backups to a floppy? (Y/N) N
Had to enter machine name/password (lucy4/lucy4).
Used defaults until you get incredibly boring musiac.
Zipping.
Must have winzip installed. "Start" "Programs" "WinZip" [winzip 7.0]
In "Winzip" window, "New", and in "New Archive" pop-up enter your filename.
"Add" window will pop-up, select "Add sub folders" and click "Add with
wildcards".
Setting Up WebSphere for Persistance
You must first configure a datasource. This is done via the WS Admin Console. From the "Types" tab, select "JDBC Drivers", right click and select "Create". A "Create a JDBC Driver" pop-up will appear, enter a Name for your driver and dependant upon what database you use, select the appropriate "Implementation class" (i.e. for DB2, select "com.ibm.db2.jdbc.app.DB2Driver"). For the remaining fields, use the defaults, and click "Create.
You must then create your Datasource. From the "Types" tab, select "Datasources", right click and select "Create". A "Create a Datasource" pop-up will appear, enter a Name for your datasource, and the database name for your database. For the database you can either use "was", or alternatively create your own database and enter it's name in the database name field. For the "Driver", select the driver you just created. For the remaining fields, use the defaults, and click "Create".
You must now configure your Session Manager's datasource. From the "Topology" tab, scope down to your "Node", application server (i.e. "Default Server"), "servletEngine", and select "Session Manager". In the "Enable" tab, select "Enable Persistent Sessions:" "Yes" radio button. In the "Persistance" tab select "directodb", and in the "JDBC Data Source" section enter you datasource name (as per above), and the associated Userid/Password. Click the "Apply" button, restart your application server, and start serving.
Upon the invocation of the first servlet/jsp that is executed (that
uses persistant store), the database table "sessions" will be created and
is viewable in your database GUI (i.e. "DB2 Control Center", yournodename,
"Instances", "DB2", "Databases", "WAS", "Tables", as "SESSIONS".)
You are now persistantly storing.
Setting
Up WebSphere With *.jsp's and beans.
Classpath's and File Placement Options
When setting up individual applications within WebSphere, you must consider
your environment and design your architecture accordingly. For example,
if you wish to have beans shared between applications, you should define
their
classpath in the application server's "Command Line Arguments" (see
WebSphere
CLASSPATHs for additional details). However, if you wish
to have individual beans used only by the associated application, define
their classpath in the applications class path. BUT, if 2 applications,
within the same application server, have a bean defined in each of the
applications which has the same name, you will experience contention in
the form of an exception.
Normally, the first time you execute a jsp file, WebSphere compiles
this into a servlet source code file (i.e "yourservlet.java"), and subsequent
compiles this into an actual Java class code. You have the option
to keep this interim Java code by using the WebSphere console, selecting
your particular JSP processor, and in the "Advanced" tab, in "Init Parameters:",
including the value "keepgenerated" in the "Init Parm Name", and the value
of "true" in the "Init Parm Value" fields.
All files in Web Server
In the WS Admin console set the "default app" "Advanced" parameter "Document Root" to "c:\IBM HTTPServer\htdocs", and include this same path in the "Classpath", "Apply" it, and restart the app server (select the appserver, right mouse and select "Restart Web App". Make sure you have the correct version of the jsp processor. The later version (1.01), the version shows up in the web app jsp name.
To change jps processors, you must delete the old one in the "Topography" view, and add the new one in the "Tasks" view (otherwise you will have to enter a path name).
In the "htdocs" directory, place your *.jsp's, *.html's, and any bean code. From "htdocs", compile your beans using the "javac -d ../htdocs yourbean.java" which will take care of any "package" structure, and place the *.class files in the appropriate directory.
From a web browser, request your jsp (i.e., "http://shayden/test1.jsp"),
and you should see it displayed.
All Files in Default App
In the WS Admin console set the "default app" "Advanced" parameter "Document Root" to "C:\WebSphere\AppServer\hosts\default_host\default_app\web". If the *.jsp files requires any classes ( i.e. beans), include "C:\WebSphere\AppServer\hosts\default_host\default_app\servlets" (use the appropriate slashes for your platform) path in the "Classpath", "Apply" it, and restart the app server (select the appserver, right mouse and select "Restart Web App". Make sure you have the correct version of the jsp processor. The later version (1.01), the version shows up in the web app jsp name.
To change jps processors, you must delete the old one in the "Topography" view, and add the new one in the "Tasks" view (otherwise you will have to enter a path name).
In the "C:\WebSphere\AppServer\hosts\default_host\default_app\web" directory,
place your *.jsp's, *.html's, etc.. Place your java code in the appropriate
directory (i.e., "C:/WebSphere/AppServer/hosts/default_host/default_app/servlets"
(or subdirectories, e.g. "com/ibm/yadayadayada") compile your beans using
the "javac -d ../web yourbean.java" which will take care of any "package"
structure, and will place the *.class files in the appropriate directory.
From a web browser, request your jsp (i.e., "http://shayden/test1.jsp"),
and you should see it displayed.
Problems
If you get an Error 404 on the browser, your document root is probably
wrong or you put the files in the wrong directory. If you get an
Error 500, (i.e., "Unable to load class com.painewebber.ces.tester"), you
probably have your classes screwed up or in the wrong place.
Setting
Up Multiple Application Servers To Share Session Data
There are actually two main methods you can use to create your application server. The first method describes the use of an "automated" way which creates most of the required features needed for you application. A "manual" method will then be described with which you must manually create your required features.
To trace all session activity, the trace specification should include com.ibm.ejs.dbm, com.ibm.servlet.session, com.ibm.servlet.personalization.sessiontracking (all with ~.*=all=enabled).
Automatic Method
From the "Tasks" tab, select "Configure an Application Server", and click the "Start" button. In the "Task Wizard Instructions" frame, under the "Application Server Properties", select the "General" tab, and in this section enter the name of your application server in the "Application Server Name" field, then click the "Next" button. In the next window select whether to start the server after creating it, then click "Next". In the next window select the node your application server will run on (i.e. "shayden"), and click "Next". In the next window add/remove any Enterprise Beans and click "Next". In the next window you must enter information related to your datasource. In the "General" tab, enter a container name (or accept the default), click "Next". In the "Virtual Host Name" panel, enter a name or select by scoping down to your host name and selecting it, click "Next". In the "Servlet Engine" screen either use the default name or enter your own engine name, then click "Next". In the "Web Applications" screen, either use the default name or enter your own selection. In the "Advanced" tab of the Web Aplications screen you can specify your "Document Root" and "Classpath" values or accept the defaults, click "Next". In the "System Servlets" screen, you can select whether to "Enable File Servlets", "Serve Servlets by Classname", and which version of JSP to use, then click "Finished". Note that at any time before you click "Finished", you can go forward and backward to modify your inputs.
In the normal example the Document Root will be the "web" directory,
and the Classpath will be the "servlet" directory. The "web" will
contain any html or jsp files, and the "servlet" will contain any java
classes/jars.
(OR)
Manual Method
From the "Topography" tab, scope down to your node (i.e. "shayden"), select the node, right click, and select "Create" , and click on "Application Server". In the resulting "Application Server Properties" pop-up, under the "General" tab, enter the name of your application, accept all other defaults and click "Create". In the rusulting pop-up, in the "General" tab, enter your servlet engine name and click "Create". Back in the Topology tab, select your new engine, right click on it and select "Create" and click on "WebApplication". In the resulting pop-up, in the "General" tab, enter a name for your application and a "Web Application Web Path" (this is what you will type in the url of a browser to get servlets/jsps from your application). In the "Advanced" tab, enter your web name. While you are entering the Web Application Web Path, the "Classpath" and "Document Root" fields were automatically being filled in under the "Advanced" tab. You can change these values if you wish. In the normal example the Document Root will be the "web" directory, and the Classpath will be the "servlet" directory. The "web" will contain any html or jsp files, and the "servlet" will contain any java classes/jars.
To be able to serve jsp pages you must also add a JSP enabler. In the "Tasks" tab of the Admin Console, select "Add a JSP enabler" and click the "Start" icon. Select Web Application and the appropriate version of the JSP Enabler and click "Finished". In the "Topology" tab, scope down to your new application, servlet engine, and web app and select the web app.
To use the Error Reporter servlet, right mouse click on your application, select "Create", and click on "Servlet". In the resulting pop-up, enter your servlet name (i.e. "Error Reporting Facility"), select the "Web Application", and enter the "Servlet Class Name" (i.e."com.ibm.servlet.engine.webapp.DefaultErrorReporter"). In the "Servlet Web Path List", click the "Add" button and in the resulting pop-up, enter the appropriate web path (i.e. "/webapp/SteveApp1WebApp/ErrorReporter") and click "OK". Then click "Create" and your servlet should show up in the topology.
To use the File Serving Enabler servlet, right mouse click on your application,
select "Create", and click on "Servlet". In the resulting pop-up,
enter your servlet name (i.e. "File Serving Enabler"), select the "Web
Application", and enter the "Servlet Class Name" (i.e."com.ibm.servlet.engine.webapp.SimpleFileServlet").
In the "Servlet Web Path List", click the "Add" button and in the resulting
pop-up, enter the appropriate web path (i.e. "/webapp/SteveApp1WebApp/")
and click "OK". Then click "Create" and your servlet should show up in
the topology.
The following data is related to e-fix's only. Some of these e-fixes
will have been incorporated in subsequent PTF's. It is YOUR job to see
if this is the case and whether you can tell a customer to go to the latest
PTF with assurance the e-fix is in.
Clones Cannot
Share Session Data in 3.5
100
Percent CPU Utilization Until First Servlet Called In 3.02x
Clones Cannot Share Session Data in 3.5
Problem: Customer configuration uses 2 clones using sessions. The first clone sets ("putValue") some data. A second servlet on a second clone does a get session but cannot get the data (i.e. "getValue").
Analysis: Turn on session tracing (trace spec = com.ibm.servlet.personalization.sessiontracking.*=all=enabled). In the resulting session trace log, at the point the associated servlet engine is started, if you see a log "ServletEngine A Servlet Specification Level: 2.2 - WebSphere 3.5 Compatibility Mode", the servlet engine is running in the old mode. If this is the case, later in the logs you should see where the second servlet tries to get a value, you will see a "selectAndLock - right before rs.next() call following a select for update of select id, propid, appname, listenercnt, lastaccess, creationtime, maxinactivetime, username, small, medium, large from sessions where id = ? and propid = ? and appname = ? for update". In this instance the "appname" is a signal that there is a defect.
Resolution: Install e-fix "pq44646" from "wasdoc0" APARS
directory.
100 Percent CPU Utilization Until First Servlet Called In 3.02x
Problem: Upon app server startup, CPU usage spikes to a large percentage until the first servlet is server, afterwards dropping to normal usage.
Analysis: Phenomena can be observed by using "xload" or "monitor -a" on Unix platforms, or "Start", "Programs" "Administrative Tools", "Performance Monitor" on NT. In P "Performance Monitor", select "Edit"" and click on "Add to Chart". Add "% Processor Time" and "% User Time". Startup your application server and watch the mips fly until you hit the servlet.
Solution: Apply e-fix "pq42764" from "wasdoc0" "APARS"
directory.